Deephaven Residential Mortgage Trust 2022-1 Deephaven

Presale:

Deephaven Residential Mortgage Trust 2022-1

January 24, 2022

Preliminary Ratings

Class

Preliminary

rating(i) Class type

Initial interest

rate (%)(ii)

Preliminary amount

($)

Credit enhancement

(%)(iii)

A-1 AAA (sf) Senior Fixed 262,501,000 29.20

A-2 AA (sf) Senior Fixed 17,797,000 24.40

A-3 A (sf) Senior Fixed 27,437,000 17.00

M-1 BBB (sf) Mezzanine Fixed 20,762,000 11.40

B-1 BB (sf) Subordinate Net WAC 16,314,000 7.00

B-2 B- (sf) Subordinate Net WAC 16,128,000 2.65

B-3 NR Subordinate Net WAC 9,826,187 0.00

XS NR Monthly excess cash

flow

(iv) Notional(v) N/A

A-IO-S NR Excess servicing (vi) Notional(v) N/A

R NR Residual N/A N/A N/A

Note: This presale report is based on information as of Jan. 24, 2022. (i)The ratings shown are preliminary. This report does not constitute a

recommendation to buy, hold, or sell securities. Subsequent information may result in the assignment of final ratings that differ from the

preliminary ratings. (ii)Interest can be deferred on the classes. Fixed coupons are subject to the pool's net WAC. The coupon on class B-1, B-2,

and B-3 equals net WAC of the pool. (iii)This credit enhancement is solely from subordination. Excess spread also provides credit enhancement.

(iv)Excess of the net WAC rate minus the coupon on the class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 notes in addition to any prepayment

premiums. (v)Notional amount equals the loans' aggregate stated principal balance. (vi)Excess servicing strip minus compensating interest and

advances owed to the servicer. WAC--Weighted average coupon. NR--Not rated. N/A--Not applicable.

Profile

Expected closing date Jan. 28, 2022.

Cut-off date Jan. 1, 2022.

First payment date Feb. 25, 2022.

Final scheduled

payment date

Jan. 25, 2067.

Notes' amount,

including unrated

classes

$370.77 million in aggregate.

Presale:

Deephaven Residential Mortgage Trust 2022-1

January 24, 2022

PRIMARY CREDIT ANALYST

Manmadh K Venkatesan, CFA

Toronto

+ 1 (212) 438 4569

manmadh.balaji

@spglobal.com

SECONDARY CONTACTS

Kalpesh S Ghule

Centennial

+ 1 (303) 721 4157

kalpesh.ghule

@spglobal.com

Diane Lebowitz

New York

+ 212-438-1524

diane.lebowitz

@spglobal.com

Kimball Ng

New York

+1 212-438-2250

kimball.ng

@spglobal.com

Jessica Barbara

New York

+ 1 (212) 438 1726

jessica.barbara

@spglobal.com

SURVEILLANCE CREDIT ANALYST

Truc T Bui

San Francisco

+ 1 (415) 371 5065

truc.bui

@spglobal.com

See complete contact list at end of article.

www.spglobal.com January 24, 2022 1

on the last page.

2784494

Profile (cont.)

Collateral type First-lien, fixed- and adjustable-rate mortgage loans secured by single-family residences,

planned-unit developments, condominiums, two- to four-family homes, and one townhouse. The

pool consists of 752 loans backed by 829 properties that are primarily non-qualified mortgage

loans and ability-to-repay exempt loans; of the 752 loans, 17 are cross collateralized, which were

broken down to their constituents at the property level (making up 94 properties).

Credit enhancement For each class of rated notes, subordination of the notes that are lower in the payment priority

and excess spread.

Participants

Issuer Deephaven Residential Mortgage Trust 2022-1.

Aggregator Deephaven Mortgage LLC.

Servicing administrator Pretium Residential Credit Management LLC.

Seller RCF II Loan LLC.

Sponsor and advance reimbursement

party

RCF II Loan Acquisition L.P.

Depositor, administrator RCF II Master Depositor LLC.

Master servicer Computershare Trust Co. N.A.

Paying agent, certificate registrar,

note registrar, and REMIC

administrator

Computershare Trust Co. N.A.

Servicers Selene Finance L.P. and Shellpoint Mortgage Servicing.

Indenture trustee Wilmington Trust N.A.

Owner trustee Wilmington Savings Fund Society FSB.

Custodian U.S. Bank N.A.

Originators ACC Mortgage, contributing 27.00% of the pool by balance; Deephaven Mortgage

LLC, contributing 22.64%; and various other originators, contributing 50.36%, each

of which make up less than 10.00% of the collateral.

Initial purchasers Goldman Sachs & Co. LLC, Barclays Capital Inc., and Credit Suisse Securities (USA)

LLC.

REMIC--Real estate mortgage investment conduit.

Primary Originators Making Up More Than 10.0% Of The Collateral

Originator By balance (%) Due diligence (%) Originator ranking

ACC Mortgage 27.00 100 N/A

Deephaven Mortgage LLC 22.64 100 AVERAGE

Top five originators 71.83 100 N/A

Top 10 originators 88.32 100 N/A

N/A--Not applicable.

www.spglobal.com January 24, 2022 2

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Servicers

By balance (%) S&P Global Ratings' select servicer Operation

Selene Finance L.P. 88.09 Yes Primary servicer

Shellpoint Mortgage Servicing 11.91 Yes Primary servicer

Rationale

The preliminary ratings assigned to Deephaven Residential Mortgage Trust 2022-1's (DRMT

2022-1) mortgage-backed notes reflect our view of:

- The pool's collateral composition (see the Collateral Summary section below);

- The credit enhancement provided for this transaction;

- The transaction's associated structural mechanics;

- The mortgage aggregator, Deephaven Mortgage LLC (Deephaven);

- The transaction's representation and warranty (R&W) framework;

- The geographic concentration;

- The 100% due diligence results consistent with represented loan characteristics; and

- The impact the COVID-19 pandemic will likely have on the performance of the mortgage

borrowers in the pool and liquidity available in the transaction.

Environmental, Social, And Governance (ESG) Factors

Our rating analysis considers a transaction's potential exposure to ESG credit factors. For

residential mortgage-backed securities (RMBS), we view the exposure to environmental credit

factors as average, social credit factors as above average, and governance credit factors as below

average (see "ESG Industry Report Card: Residential Mortgage-Backed Securities," published

March 31, 2021). In our view, the transaction's exposure to social and environmental credit factors

is in line with the sector benchmark.

For RMBS, we generally consider social credit factors as above average because housing is viewed

as one of the most basic human needs, and conduct risk presents a direct social exposure for

lenders and servicers because regulators are increasingly focused on ensuring fair treatment of

borrowers. Social risk is generally factored into our base-case assumptions for RMBS

transactions based on our consideration of the origination platform, the R&W framework and the

third-party due diligence that informs our view of credit underwriting and compliance with

applicable consumer protections. With respect to environmental factors, our assumptions

consider physical climate risks such as floods, storms, or wildfires, which could severely damage

properties, reduce their value, and hurt recoveries if borrowers default. The transaction has a

geographically well-diversified portfolio, which reduces exposure to extreme weather events, in

our view.

The transaction's governance risk exposure is higher than our benchmark due to certain

weaknesses related to the R&Ws framework as described further below. By applying certain R&W

pool-level adjustment factors to the transaction, we have accounted for risk related to ESG credit

factors. Certain other features also provide mitigants to the transaction's governance risk

www.spglobal.com January 24, 2022 3

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

exposures including the fact that 100% of the loans in the pool were subject to a third-party due

diligence review with limited material findings (see the Third-Party Due Diligence section below for

more detail).

Overview

DRMT 2022-1 is the 16th non-qualified mortgage (non-QM) RMBS transaction from aggregator

Deephaven rated by S&P Global Ratings. Deephaven originated or purchased the mortgage loans

from 25 originators with Selene Finance L.P. (Selene) and Shellpoint Mortgage Servicing as

servicers. Deephaven has one of the longer track records of acquiring non-QM mortgage loans and

was acquired by Pretium Partners LLC in September 2019.

Noteworthy Features

High concentration of alternative and other documentation loans

The transaction includes 262 loans (26.90% by pool balance) that are property-focused investor

loans underwritten to an investment property business-purpose program that did not consider the

borrowers' income or employment in the underwriting process, as well as 384 loans (56.91% by

balance) that were verified using alternative documentation, such as personal or business bank

statements or P&L statements. We view income verification using other and alternative

documents to be a weaker standard than "full" documentation of income and, consequently,

increased our loss coverages for these loans by applying an adjustment to the foreclosure

frequencies.

Cross-collateralized loans

Approximately 5.41% of the DRMT 2022-1 pool balance consists of cross-collateralized loans with

17 loans secured by 94 properties. Cross-collateralized loans aggregate multiple properties under

one loan. These loans have full recourse to the borrower (although the enforceability of recourse

may be limited by state and local law), and all are debt service coverage ratio (DSCR) investor

loans. The maximum number of properties securing any one loan is 10 properties. We analyzed the

cross-collateralized loans as separate properties, and the issuer allocated the balance of each

loan in proportion to each property's appraisal value. We treated each property within each

cross-collateralized loan as having the same DSCR, which is calculated by aggregating the

qualifying rental income and expenses for all properties securing the loan. While our LEVELS

model does not treat cross-collateralized loans differently than loans backed by a single property,

any fixed costs associated with foreclosing on a property is duplicated for each property within the

loan. This results in a higher loss severity when the analysis is performed at the property level.

All of the cross-collateralized loans contain release provisions (per the originator guidelines)

where the borrower would need to pay 120% of the individual property loan amount in order to

remove it from the cross-collateralized loan. This predetermined payoff premium reduces the

incentive for borrowers to release the liens on stronger properties and leave only the relatively

weaker properties to back their loans.

www.spglobal.com January 24, 2022 4

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

No loans are in forbearance at closing

On March 27, 2020, the CARES Act enacted COVID-19-related relief for borrowers with

government-backed mortgage loans in the form of a temporary forbearance of up to 12 months of

scheduled payments. While nonagency loans do not fall under the CARES Act as it relates to this

forbearance, servicers have been granting forbearance plans to nonagency borrowers as well,

typically with some variations to those of the CARES Act (e.g., timeframe and approval

requirements). The updates we made on April 17, 2020, to our mortgage outlook and

corresponding archetypal foreclosure frequency (FF) levels (see "Guidance Methodology And

Assumptions For Rating U.S. RMBS Issued 2009 And Later," published April 17, 2020) account for

a portion of the borrowers entering COVID-19-related temporary forbearance plans and their

impact to the overall credit quality of collateralized pools. To the extent a securitization pool

exhibits growth levels in forbearance over time beyond those otherwise expected, additional

adjustments may be applied.

As of the Jan. 1, 2022, cut-off date, none of the mortgage loans in the pool had borrowers who had

requested or entered into a COVID-19-related forbearance plan or exited from a prior forbearance

plan.

While we recognize that temporary forbearance related to the COVID-19 pandemic could be

granted at some level in the pool going forward, we decided not to apply an additional pool-level

loss adjustment factor because, as of Jan. 13, 2022, none of the mortgage loans are currently in

an active forbearance plan. We will also continue to monitor macroeconomic and housing

conditions and update our mortgage market outlook and associated archetypal foreclosure

frequencies, as applicable.

Collateral Summary

DRMT 2022-1's assets consist primarily of newly originated fixed- and adjustable-rate, and

interest-only (IO) non-QM and ability-to-repay (ATR) exempt mortgage loans secured by first liens.

The mortgage pool consists of 752 mortgage loans with a principal balance of approximately

$370.77 million as of the cut-off date. Out of the 752 loans, 17 are cross collateralized and backed

by 94 properties.

The collateral pool, from a credit perspective, is weaker than the S&P Global Ratings' archetypal

prime pool, but is generally in line with our expectations for a nonprime mortgage residential

mortgage pool. The pool's 'AAA' loss coverage requirement was determined to be 30.20%. In our

analysis, we considered the following mortgage loan characteristics of the pool to be weaker than

the archetypal prime pool:

- Loans underwritten to alternative income documentation and DSCR;

- FICO and loan-to-value (LTV);

- Property types (condominiums and multi-family properties);

- Loan types, including adjustable-rate mortgage (ARM) and IO loans;

- Cash-out refinance loans;

- Self-employed borrowers for certain loans;

- Loan term longer than 30 years;

- Loans to borrowers who are either foreign national or non-permanent alien resident of U.S.; and

www.spglobal.com January 24, 2022 5

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

- Primarily non-QM or ATR exempt loans.

The mortgage loans consist of fixed-rate (96.56% by balance, some with IO periods) and hybrid

ARM loans (3.44% by balance, some with IO periods). These loans are fully amortizing with 15-,

25-, 30- and 40-year original terms to maturity. The weighted average seasoning for the pool is

approximately three months.

The weighted average used FICO score for the collateral pool is 729, which includes certain S&P

Global Ratings' assumptions (see table 1 for a breakdown of the pool by the borrowers' FICO

score). In the pool, there are 33 loans to foreign national and nonpermanent resident alien

borrowers (3.22% by balance), one of which are to borrowers without a FICO score. We assessed

the loan in our credit analysis using a FICO score of 678, which is approximately the mortgage

pool's average original FICO score minus one standard deviation. The weighted average FICO of

these 33 loans is 712. We applied a 1.5x multiple to the foreclosure frequencies of these 33 foreign

borrower loans.

Used credit score statistics

Table 1

Used Credit Score Statistics

FICO score Current balance (%)

750 and above 36.58

725-749 19.12

700-724 18.67

675–699 (678) 11.08

650–674 8.60

625–649 4.90

600-624 0.54

575-599 0.51

550-574 -

Below 550 -

Total 100.00

(i)Includes our assumption of a 678 FICO score for one borrower who lacks a FICO score.

Mortgage loans backed by properties that are primary residences make up approximately 65.21%

of the pool by balance. The mortgage loans are secured by first liens on single-family residences

including townhouse (59.50% by balance), planned-unit developments (PUDs) (24.15%), two- to

four-family homes (9.34%), and condominiums (7.01%).

As compared to the most recently rated DRMT 2021-4 transactions, the collateral characteristics

of DRMT 2022-1 are slighter weaker, primarily due to a higher proportion of loans underwritten to

alternative income documentation and higher proportion of cash-out refinance loans. The loans

are also less seasoned. On the positive side, the collateral pool of DRMT 2022-1 does not include

any loans that were subject to COVID-19-related forbearance plan and fewer loans with

credit-related delinquencies, as well as fewer foreign borrowers.

www.spglobal.com January 24, 2022 6

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Chart 1

Table 2

Collateral Characteristics

DRMT

2022-1

DRMT

2021-4

DRMT

2021-3

DRMT

2021-2

DRMT

2021-1

DRMT

2020-1

Archetypal

pool(i)

Closing pool balance (mil. $) 370.8 383.2 379.8 203.3 146.2 407.9 N/A

Closing loan count (no.) 752 751 857 452 395 959 N/A

Avg. loan balance ($) 493,039 510,217 443,227 449,857 370,036 425,380 N/A

WA original CLTV (%)(ii) 74.0 73.3 74.7 71.3 75.6 74.3 75.0

WA current CLTV (%) 73.8 72.9 71.3 70.9 74.9 74.1 75.0

WA FICO(iii) 729 729 729 731 702 701 725

WA current rate (%) 4.9 5.1 5.6 5.8 6.6 7.2 N/A

WA original term (mos.) 375 376 368 373 360 365 360

WA seasoning (mos.) 3 3 12 4 9 4 0-6

WA debt-to-income (%) 32.6 32.5 34.0 34.6 34.2 36.3 36.0

WA DSCR (non-zero) 1.4 1.5 1.3 1.0 1.1 1.1 N/A

Owner occupied (%) 65.2 56.4 72.5 46.1 54.2 78.1 100.0

Single-family (including

unattached and attached PUD) (%)

83.7 78.5 85.6 79.4 88.7 88.4 100.0

Five- to 10-family homes/mixed

use/condotel (%)

0.0 0.4 0.4 1.6 1.0 0.0 0.0

Adjustable-rate loans (%) 3.4 12.6 19.5 8.7 40.1 46.3 0.0

Loans with IO payments (%) 16.8 23.8 16.2 16.7 10.1 11.6 0.0

www.spglobal.com January 24, 2022 7

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Table 2

Collateral Characteristics (cont.)

DRMT

2022-1

DRMT

2021-4

DRMT

2021-3

DRMT

2021-2

DRMT

2021-1

DRMT

2020-1

Archetypal

pool(i)

Purchase (%) 55.8 57.5 67.1 54.4 55.5 57.6 100.0

Cash-out refinancing (%) 32.8 20.7 19.5 25.7 36.8 33.2 0.0

Full documentation (%) 13.5 16.8 24.2 21.1 13.8 31.4 100.0

Alternative/bank statement

documentation (%)

56.9 48.7 53.6 31.6 42.2 53.7 0.0

Other/asset depletion/DSCR

documentation (%)

29.6 34.5 22.2 47.2 44.0 14.9 0.0

Self-employed borrowers (%) (iv) 62.9 56.1 63.8 37.4 49.0 62.8 0.0

Loans with co-borrowers (%) 30.1 20.2 22.8 29.5 23.6 33.8 0.0

Loans to borrowers with multiple

mortgages (%)(v)

9.6 6.8 8.7 6.3 10.2 2.6 N/A

Loans to foreign borrowers

(%)(foreign national and

non-permanent resident aliens)

3.2 7.6 11.8 3.7 4.1 2.3 0.0

Modified loans (%)(vi) 0.0 0.0 0.8 0.0 4.8 0.0 0.0

PCEs (%)(vi) 1.5 0.1 0.9 0.2 1.2 4.3 0.0

Current (%) (vii) 99.3 100.0 98.2 98.6 93.2 100.0 100.0

30+ day delinquent (%) 0.7(viii) 0.0 1.8 1.4 6.8 0.0 0.0

Length of P&I advancing (mos.)(ix) 6 6 6 6 6 6 Full

Pool-level adjustments (multiplicative factors)

Geographic concentration 1.00 1.00 1.00 1.01 1.00 1.00 1.00

Mortgage operational

assessment

1.00 1.00 1.00 1.00 1.00 1.00 1.00

Representations and

warranties

1.10 1.10 1.10 1.10 1.10 1.10 1.00

Other (i.e., loan

modification/PCE/due

diligence)

1.02 1.00 1.01 1.01 1.02 1.06 1.00

Loans in forbearance/deferred

payments related to COVID-19

1.00 1.00 1.03 1.02 1.15 N/A N/A

Combined pool-level

adjustments(x)

1.12 1.10 1.14 1.14 1.29 1.17 1.00

Loss estimation

'AAA' loss coverage (%) 30.20 29.65 25.05 28.80 43.45 32.10 7.50

'AAA' foreclosure frequency (%) 53.78 54.88 47.91 56.52 74.90 56.68 15.00

'AAA' loss severity (%) 56.15 54.03 52.29 50.96 58.01 56.63 50.00

'BBB' loss coverage (%) 10.60 10.45 8.80 10.50 17.10 10.95 1.92

'BBB' foreclosure frequency

(%)

29.89 30.76 26.86 32.27 46.55 29.67 6.41

'BBB' loss severity (%) 35.46 33.97 32.76 32.54 36.73 36.91 30.00

www.spglobal.com January 24, 2022 8

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Table 2

Collateral Characteristics (cont.)

DRMT

2022-1

DRMT

2021-4

DRMT

2021-3

DRMT

2021-2

DRMT

2021-1

DRMT

2020-1

Archetypal

pool(i)

'B' loss coverage (%) 3.30 3.30 2.90 3.80 7.25 2.90 0.65

'B' foreclosure frequency (%) 12.41 13.07 11.72 14.67 25.48 10.33 3.25

'B' loss severity (%) 26.59 25.25 24.74 25.90 28.45 28.07 20.00

(i)As defined in our Feb. 22, 2018, criteria article. The guidance document published April 17, 2020, reflects a revision to our 'B' (base-case)

projected foreclosure frequency assumption for an archetypal loan to 3.25% from 2.50%. (ii)CLTV calculation includes second lien/ junior

mortgages identified. (iii)FICO reflects the most recent scores obtained with certain analytical assumptions. (iv)Self-employed considered in

our analysis. (v)Limited to borrowers who have multiple mortgage loans or properties included in the securitized pool. (vi)Limited to modified

and PCE loans considered in our analysis. (vii)Loans in forbearance are treated as current and included in the model forbearance adjustment.

This pool does not contain any loans that are in forbearance. (viii)Loans that are 30 days delinquent as of cut-off date (ix)Months of P&I

advancing on a delinquent mortgage loan to the extent such advances are deemed recoverable. (x)Combined pool-level adjustments are the

product of each pool-level adjustment listed above. WA--Weighted average. CLTV--Combined loan-to-value ratio. DSCR--debt service

coverage ratio. PUD--planned-unit development. IO--Interest only. PCE--Prior credit event. P&I--Principal and interest. N/A--Not applicable.

www.spglobal.com January 24, 2022 9

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Chart 2

On the transaction closing date, the affiliate of the sponsor will deposit the portion of the

mortgage loans it acquired from the aggregator to the seller (RCF II Loan LLC), and the remainder

of the loans are acquired by the seller directly. The transaction is structured as a true sale of all

the mortgage loans from the seller to the depositor (RCF II Master Depositor LLC), and then a true

sale from the depositor to the issuing trust (DRMT 2022-1) on the closing date. The issuing trust

transfers the notes to the depositor. The depositor sells the offered notes to the initial purchasers,

which will sell them to third-party investors. The depositor sells the non-offered notes, as well as

the notes required to be held to satisfy the risk retention rules, to the sponsor or affiliate.

In rating this transaction, S&P Global Ratings will review the legal matters it believes are relevant

to its analysis, as outlined in its criteria.

www.spglobal.com January 24, 2022 10

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Strengths And Weaknesses

We believe the following characteristics strengthen the DRMT 2022-1 transaction:

- Deephaven's overall mortgage operational assessment (MOA) ranking on its acquisition

process for non-QM mortgage loans, which accounts for, among others, the company's highly

experienced management team and mature aggregation platform relative to other non-QM

peers is AVERAGE.

- The third-party due diligence providers--Selene New Diligence Advisors LLC (NDA) and Incenter

LLC d/b/a Edgemac, both of which are on our list of reviewed providers--combined performed

due diligence on 100% of the pool's loans. Their review encompassed regulatory compliance,

credit (underwriting) compliance, property valuations, and data quality.

- The senior classes benefit from a credit support floor where no principal is paid to the

subordinate classes until the class A notes are retired.

We believe the following factors weaken the DRMT 2022-1 transaction:

- The mortgage pool includes property-focused investor loans that were underwritten to an

investment property business purpose program (26.9% by balance). The DSCR of these loans

range from 0.42 to 3.2, and the weighted average DSCR for these loans was 1.36. We applied an

adjustment factor ranging from 3.15x to 6.00x to the foreclosure frequencies for these loans.

- Income on certain mortgage loans (56.91% by balance) was verified using "alternative" income

documentation (personal/business bank statements or P&L statements). All of such loans were

underwritten to self-employed borrowers. We view income verification using alternative

documentation to be a weaker standard than "full" documentation of income. Consequently,

we increased our loss coverages for these loans by applying an adjustment to the foreclosure

frequencies. We applied an adjustment factor of 2.00x and 1.75x to the foreclosure frequencies

for loans using 12-23 months of bank statements or P&L statements, and at least 24 months of

bank statements, respectively.

- About 16.36% of the loans are backed by property types other than single-family homes and

PUDs, such as condominiums, and two- to four-family houses properties. We applied an

adjustment to their foreclosure frequencies or loss severity to account for this risk.

- Approximately 62.89% of the pool were loans made to self-employed borrowers. We applied a

1.10x adjustment factor to our loss estimates for these loans. The self-employment adjustment

factor is not applicable to property-focused investor loans.

- Non-QM loans, which have an increased risk of ATR challenges and associated loan losses,

comprise 68.56% of the pool. We applied an adjustment to loss severities per our QM criteria to

account for this risk.

- Some loans (16.80% by pool balance) in the pool include an IO period, including 68 fixed-IO

loans (13.36% by pool balance) and 13 ARM-IO loans (3.44% by pool balance). We increased the

foreclosure frequencies and loss severity of these loans per our criteria to account for this risk.

- The loan purpose for 267 loans (32.83% by pool balance) is a cash-out refinance, with an

average cash-out amount of $163,054; 63 loans have cash-out amounts greater than or equal

to $200,000.

- Some mortgage loans (25.63% of the pool by balance) were made to borrowers with current

FICO scores below 700 (including our assumption of a 678 FICO score for one borrower who lack

www.spglobal.com January 24, 2022 11

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

a FICO score). The mortgage pool's loss estimate has been increased to account for the

increased default risk of these loans.

- The mortgage loan aggregator (Deephaven) and sponsor, both unrated entities, make the R&Ws

for this transaction. We have not reviewed the representations made by the originators because

those were not assigned to the trust. The R&W framework is weak because the testing of any

breaches (other than any loans showing ATR-related losses), is at the controlling holder's (the

majority owner of class XS and initially, either the sponsor or an affiliate of the sponsor) option.

In addition, the early payment default (EPD) covenant the sponsor provided is weaker than

typically seen in other non-QM transactions. The weaknesses of the framework are somewhat

mitigated by third-party due diligence on 100% of the loans, along with the alignment of

interest between noteholders and the sponsor, which holds the first-loss piece and retains risk

via a 5% slice (horizontal) of the capital structure. Consequently, we applied an R&W factor of

1.10x, which increased our loss expectations at all ratings by 10.0%.

Credit Analysis And Assumptions

Our analysis of the DRMT 2022-1 collateral pool considered a number of factors, including certain

loan-level characteristics. The details of our analysis are described below.

Property cash flow underwriting

We considered the underwriting methods employed for investor focus business purpose loans

(26.9% by pool balance), given that the qualifying metrics do not use traditional borrower

characteristics such as personal income and liabilities, but instead rely on the property's

propensity to generate cash flow from tenants. These loans are underwritten using a DSCR

generally calculated as rental income divided by mortgage payment liability (including taxes and

insurance). The weighted average DSCR of the pool is 1.36x. There are no loans in the pool that

were qualified without using this ratio based on other characteristics, such as FICO scores and

LTV ratios.

We consider the strength of the DSCR and apply adjustment factors to the foreclosure frequency

with higher factors applied to lower DSCRs and lower factors to higher DSCRs. This factor ranges

from 3.15x to 6.00x. The low end of the range (3.15x) was calibrated such that a DSCR loan with a

high DSCR (i.e., greater than or equal to 1.27x) is treated similarly to a weak traditionally

underwritten investor property (i.e., underwritten to the borrower's income) with less than 12

months of income verification and poor debt-to-income (DTI) attributes (i.e., where the maximum

adjustment factors for full-income documentation and DTI are assumed), all else being equal,

given the limited performance history of DSCR loans through an economic cycle.

The weighted average DSCR loss coverage adjustment factor for this pool was 1.69x, which we

believe adequately addresses the additional risk of DSCR loans that rely on the property cash flow

rather than personal income and liabilities.

Documentation type

The sponsor guidelines allow income verification using paystubs, W-2s/W-2 equivalents, tax

returns, written verification of income and alternative income documentation such as bank

statements, profit-and-loss (P&L) statements, and CPA letters. The sponsor also considers asset

depletion as borrower income and to qualify for monthly payments. For investment property loans

www.spglobal.com January 24, 2022 12

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

underwritten to the property's rental income, the sponsor considers the properties' DSCRs (see

table 3).

Table 3

Documentation Type (Income Verification Type/Length)

Loan

count

(no.)

Current

balance

(%)

Alternative income

verification length

(WA # of months)

Foreclosure

frequency

adjustment

factors (x)

'AAA' foreclosure

frequency without

pool adjustment

factors (%)

Full documentation

Appendix Q/qualified

mortgage

1.00

Full (24+ months) 63 9.64 1.00 29.2

Full (24+ months) WVOE only 1 0.05 1.00 42.5

Full (12-23 months) 27 3.76 1.25 35.6

Full (12-23 months) WVOE

only

1.25

Full (1-11 months) 1 0.04 1.50 73.4

Full (1-11 months) WVOE only 1.50

Alternative documentation(i)

24+ months

Business bank statements 33 4.48 24.0 1.75 50.6

Personal bank statements 6 0.68 24.0 1.75 44.7

Personal and business bank

statements

20 3.94 24.0 1.75 54.4

P&L statements 1.75

12-23 months

Business bank statements 238 36.20 12.1 2.00 50.4

Personal bank statements 21 2.35 12.0 2.00 43.1

Personal and business bank

statements

32 6.24 12.0 2.00 51.0

P&L statements 34 3.03 12.0 2.00 60.0

1-11 months

Business bank statements 2.25

Personal bank statements 2.25

Personal and business bank

statements

2.25

P&L statements 2.25

Other documentation

Other (DSCR) 262 26.90 3.15-6.00 61.5

Other (zero DSCR/no ratio) 6.00

www.spglobal.com January 24, 2022 13

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Table 3

Documentation Type (Income Verification Type/Length) (cont.)

Loan

count

(no.)

Current

balance

(%)

Alternative income

verification length

(WA # of months)

Foreclosure

frequency

adjustment

factors (x)

'AAA' foreclosure

frequency without

pool adjustment

factors (%)

Other (asset

underwriting/depletion)

14 2.69 3.00 39.1

(i)The documentation source may include other secondary documentation types. WA--Weighted average. P&L--Profit and loss. DSCR--Debt

service coverage ratio.

As mentioned above, 26.9% of the pool by balance is made up of loans underwritten to a DSCR,

based on a property's rental income cash flow. The DSCR calculations provided by the issuer for

these loans ranged from 0.42 to 3.20, and the weighted average DSCR for these loans was 1.36.

Seventeen DSCR loans are cross-collateralized, which aggregate multiple properties under one

loan and are typically made to experienced investors. The properties within a loan generally share

homogenous features, such as loan purpose (rate or cash refinance), property type, and

geographical location. The DSCR for these loans is calculated by aggregating the qualifying rental

income and expenses for all properties securing the loan. We split these 17 loans into 94

property-level constituents where appropriate for our analysis (see table 4 for the average 'AAA'

and 'B' foreclosure frequencies for the DSCR, non-DSCR, and cross-collateralized DSCR loans). All

of the cross-collateralized loans contain a release provision where the borrower would need to pay

120.00% (per the originator guidelines) of the individual property loan amount in order to remove it

from the cross collateralized loan. This predetermined payoff premium reduces the incentive for

borrowers to release the liens on stronger properties and leave the relatively weaker properties to

back their cross-collateralized loan.

Table 4

Average 'AAA' And 'B' FFs For The DSCR Loans(i)

Loan count (no.) % of balance Avg. DSCR Avg. FF (%) ('AAA') Avg. FF (%) ('B')

DSCR (non cross-collateralized)

DSCR < 1 0 0.00 0.00 0.00 0.00

1 <= DSCR <= 1.27 0 0.00 0.00 0.00 0.00

DSCR > 1.27 0 0.00 0.00 0.00 0.00

Non-DSCR 0 0.00 - 0.00 0.00

Cross-collateral DSCR(i)

DSCR < 1 - - - - -

1 <= DSCR <= 1.27 0 0.00 0.00 0.00 0.00

DSCR > 1.27 0 0.00 0.00 0.00 0.00

Non-DSCR - - - - -

(i)17 cross-collateralized loans are backed by 94 properties. Asset depletion--14 loans. FF--Foreclosure frequency. DSCR--Debt service

coverage ratio.

For 92 loans (approximately 13.49% of the pool balance), traditional (full) documentation was

used for fully verifying and calculating the borrowers' qualifying income (e.g., pay stubs, W-2s,

www.spglobal.com January 24, 2022 14

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

personal and business tax returns, IRS transcripts, and written verification of income). We applied

a documentation type adjustment factor ranging from 1.00x to 1.50x, depending on the length of

the income verification, for these loans.

We classified all loans underwritten to nontraditional sources of borrower income documentation,

such as bank statements (business or personal), CPA- or accountant-prepared letters

documenting income, and P&L statements as alternative documentation loans. Alternative

documentation was used on 384 loans (56.91% by pool balance), all of which used 12 or more

months of business/personal bank statements or P&L statements for income verification. We view

income verification using alternative documentation to be a weaker standard than full

documentation of income. As result, we increased our loss coverages for these loans by an

adjustment factor ranging from 1.75x to 2.00x, depending on the length of verification.

Fourteen loans in the pool (2.69% by pool balance) were underwritten considering accumulated

assets rather than a verified income stream. We classified these loans as "other" documentation

loans and applied a 3.00x adjustment to the foreclosure frequencies.

QM and ATR standards

The Consumer Financial Protection Bureau issued final regulations for mortgage loans with

applications submitted on or after Jan. 10, 2014, specifying the standards for a QM. Per the

designation provided by the sponsor and verified by the due-diligence firms, most loans are

categorized as non-QM/compliant (68.56% by balance) or not covered/exempt (31.44% by

balance).

Table 5

Qualified Mortgage Breakout

QM status Pool balance ($) % by pool balance Loan count (no.)

QM/non-HPML -- -- --

QM/HPML -- -- --

Non-QM/compliant 254,197,858 68.56 452

Not covered/exempt 116,567,329 31.44 377

QM--Qualified mortgage. HPML--Higher-priced mortgage loan.

Under the ATR rule (as more fully described in our QM criteria; see Appendix I of "Methodology And

Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018), the

originator and any assignee are jointly and separately liable for certain damages that may be

incurred from noncompliance with the rule. We applied our criteria for each loan subject to the

rule, which increased our loss coverage estimates at each rating category for the portion of

non-QM loans.

Servicer advancing obligations

For any loan that is not in forbearance, Selene or Shellpoint, the servicers, must advance

delinquent principal and interest (P&I) payments for any delinquent mortgage loan until that loan

is greater than 180 days' delinquent, at which time Selene or Shellpoint are no longer obligated to

advance delinquent P&I payments (limited P&I advances). If the P&I advancing party fails to

advance P&I, the master servicer, Computershare Trust Co. N.A., is obligated to make those

www.spglobal.com January 24, 2022 15

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

advances. Unlike P&I advances, the servicer must make advances of delinquent taxes and

insurance (and other property preservation advances) for any delinquent mortgage loan until the

related property is liquidated or the servicers deems the advance to be nonrecoverable. We

adjusted the loss severities in our model to account for this limited P&I advancing.

Prior credit event (PCE) classification and analysis

The borrowers on a portion of the mortgage loans have had one or more PCEs, such as

bankruptcies or housing-related PCEs (foreclosures, short-sales, deed-in-lieu of foreclosure, etc.)

that may have limited their access to loan products offered by the various agencies. We believe

that a borrower's behavior surrounding a PCE could indicate what it would do when faced with a

similar situation in the future, and suggests a greater likelihood that it would default,

notwithstanding this and other adverse performance already incorporated in its FICO score.

Therefore, these borrowers may behave similarly to borrowers who are 30-days delinquent.

We focused primarily on prior bankruptcy, foreclosure, short-sale, and deed-in-lieu events (24

months from the cut-off date for bankruptcy discharges or dismissals and 36 months from the

cut-off date for housing-related events). For loans to borrowers with more seasoned PCEs, we

believe that the associated risks associated with those PCEs are reflected in the updated FICO.

We applied a rounded pool-level PCE-related loss coverage adjustment factor of 1.02x for this

pool, which was derived from the 2.50x weighted average factor (30-day delinquent loan factor),

for 14 loans (1.45% by pool balance), and a 1.00x factor for the remaining loans in the mortgage

pool.

Geographic Concentration

S&P Global Ratings analyzes the pool's geographic concentration risk based on the

concentrations of loans in each of the core-based statistical areas (CBSAs) as defined by the U.S.

Office of Management and Budget (see Appendix II of "Methodology And Assumptions For Rating

U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018). In this transaction, the top five

CBSAs account for approximately 28.51% of the aggregate pool. Given the relative diversification

of the pool, we applied a Herfindahl adjustment factor (a concentration measure based on the

sum of the squared CBSA concentrations related to a benchmark concentration) of 1.00x to our

base loss coverage estimate.

Table 6

Geographic Concentration

CBSA code(i) CBSA State % by balance

31084 Los Angeles-Long Beach-Glendale California 10.15

38060 Phoenix-Mesa-Chandler Arizona 5.49

33124 Miami-Miami Beach-Kendall Florida 4.93

35614 New York-Jersey City-White Plains New York-New Jersey 4.14

11244 Anaheim-Santa Ana-Irvine California 3.80

Top five -- -- 28.51

(i)The CBSA code refers to the metropolitan division code, if available. CBSA--Core-based statistical area (includes metropolitan statistical

areas and metropolitan divisions where defined, as well as micropolitan statistical areas).

www.spglobal.com January 24, 2022 16

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Large Loans And Tail-Risk Considerations

As the number of loans in the transaction decreases, the effect of a single loan's losses becomes

greater. To mitigate this risk, this transaction provides for a sequential payment structure, which

does not allow for depletion, via principal payments, of the respective tranches that provide credit

support to each rated class. Due to the sequential payment mechanism, the preliminary 'AAA (sf)',

'AA (sf)', 'A (sf)', 'BBB (sf)', 'BB (sf)', and 'B- (sf)' rated classes effectively have an initial floor of

29.20%, 24.40%, 17.00%, 11.40%, 7.00%, and 2.65%, respectively.

To analyze the appropriateness of this effective credit enhancement floor, we use the approach

outlined in "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later"

published Feb. 22, 2018. Per this approach, instead of focusing on the largest loans by pool

balance at issuance, we risk-weight the loans in the transaction by focusing on those loans with

the largest expected loss exposure assuming default.

After considering the enhancement provided in the transaction and the expected pay down of the

notes, we believe the rated senior notes are sufficiently protected from tail risk as the transaction

seasons.

MOA Review

Deephaven

We conducted a MOA of Deephaven's acquisition process for non-QM mortgage loans and

assigned our overall AVERAGE MOA ranking to Deephaven. The ranking reflects our AVERAGE

qualitative subranking and AVERAGE quantitative subranking. Based on the results of our MOA,

we determined a loss coverage adjustment factor of 1.00x for Deephaven. This accounts for the

company's highly experienced management team, 100% due diligence review of its loan

acquisitions, and mature aggregation platform relative to other non-QM peers, measured against

the lack of a formal independent risk management group or internal audit function, and the fact

that its loan performance history has not experienced a housing or economic downturn.

The company was founded in June 2012 and acquired by Varde Partners Inc. in November 2014. In

September 2019, Pretium Partners LLC (Pretium) acquired Deephaven from Varde Partners LLC.

Pretium is a specialized alternative investment manager focused on real estate and corporate

credit. The company was founded in 2012 to capitalize on secular investment and lending

opportunities that arose after the global financial crisis as a result of structural changes within

the economy, the residential housing sector, and mortgage finance markets. Pretium has built an

integrated analytical and operational ecosystem within the U.S. residential housing, mortgage,

and corporate credit markets. As of Sept. 30, 2021, it manages $31.8 billion in assets, of which

over $29.0 billion is invested through either single-family rentals or residential credit.

Pretium is active in residential credit and manages a residential whole loan strategy that has

acquired loans with a total unpaid principal balance of approximately $6.1 billion through

September 2021. The company has strategically built and acquired affiliates and key partners to

optimize performance and drive its returns across residential credit. Pretium has built a vertically

integrated residential credit platform, supporting the entire lifecycle of the asset

class--origination, asset management, servicing, and disposition.

Deephaven purchases non-QM loans from preapproved correspondents and also originates loans

www.spglobal.com January 24, 2022 17

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

sourced from preapproved brokers through its wholesale channel. In 2017, Deephaven rolled out

its wholesale channel, in which the company is able to originate and fund non-QM loans that are

sourced through a network of approved brokers. It expects continued growth on the correspondent

side. The company is also licensed to originate mortgage loans in 46 states as of September 2021.

All of the company's functions take place in Charlotte, N.C.

Deephaven conducts 100% due diligence on loans that it acquires using third-party review firms

that are on S&P Global Ratings' reviewed list (see "S&P Global Ratings Publishes List Of

Third-Party Due Diligence Firms Reviewed For U.S. RMBS As Of Sept. 10, 2021," published Sept.

10, 2021). The scope of the review, which is consistent with market standards, comprises a full

re-underwrite on these loans (credit, compliance, property valuation, and fraud). All loans must be

submitted to an automated fraud and data check tool.

In addition to the third-party due diligence firm's review, Deephaven's underwriting team performs

a pre-close eligibility credit and collateral review on certain loans. Compliance reviews are

conducted by a third-party due diligence firm before the company purchases the loans.

Deephaven said it will add post-purchase quality control reviews if it starts allowing a

less-than-100% due diligence review by the third-party firm. The company also noted it has done

very few loans on a nondelegated basis. For 100% of the loans originated through the wholesale

channel, third-party due diligence firms perform post-funding compliance reviews.

Although Deephaven is one of the more mature aggregators in the non-QM sector with a longer

track record of loan performance (formed in 2012 and started purchasing loans in March 2013), it

has not yet experienced a typical (i.e., non-COVID-19 pandemic related) housing or economic

downturn. In 2021 and 2020, Deephaven's acquisitions were approximately $947 million and $580

million, respectively, up from $3.3 million in 2013. Through December 2020, Deephaven's

acquisitions totaled over $7.5 billion since it started aggregating loans. To date, Deephaven's

performance has generally been consistent with our expectations for an average non-QM

aggregator given the economic environment of the past several years.

Key assessment factors

Deephaven's strengths include:

- An experienced management team averaging over 20 years of industry experience;

- A requirement to conduct 100% due diligence (prepurchase) on all acquisitions, which includes

a full review of credit, compliance, property valuation, and fraud;

- An absence of legacy issues, allowing the company to build a holistic approach to reviewing

sellers; and

- Servicing retention on loans that it acquires.

Partly offsetting the above strengths, are, in our view, the following weaknesses:

- Its loan performance history does not cover a typical housing or economic downturn; and

- No formal independent risk management function or internal audit department.

Additionally, we conducted a primarily quantitative-focused, transaction-specific review of ACC

Mortgage, which represent 27.00% of the pool by balance. Based on our analysis, we determined

the originator's loan performance was in line with that of Deephaven's securitized portfolio.

Therefore, we concluded that the 1.00x loss coverage adjustment factor applicable for Deephaven

was also appropriate for the loans originated by ACC Mortgage.

www.spglobal.com January 24, 2022 18

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

As such, we applied an overall MOA loss coverage adjustment factor for the aggregate pool of

1.00x.

Third-Party Due Diligence Review

Selene New Diligence Advisors LLC (Selene) and Incenter LLC d/b/a Edgemac, combined,

performed third-party due diligence on 100% of the loans in the transaction. The scope

encompassed compliance, credit, and valuation reviews. After reviewing the third-party due

diligence results, we applied a final rounded adjustment of 1.00x to the loss coverage at all rating

categories.

Some loans fell within the scope of the TILA-RESPA Integrated Disclosure (TRID) rule. For these

loans, the third-party firms followed the Structured Finance Assn. RMBS 3.0 TRID Compliance

Review Scope in conducting their final loan reviews (see "Standard & Poor's Comfortable With

SFIG Draft Proposal Regarding TRID Due Diligence," published April 25, 2016). In accordance with

our criteria, we adjust our loss expectations based on our view of the firms' findings (see Appendix

III of "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published

Feb. 22, 2018).

Fourteen loans were originated by ACC Mortgage to ITIN borrowers which were reviewed by

Selene. Selene followed agency guidelines to review their legal residency status and classified

these borrowers as non-permanent resident aliens.

Highlight of the findings include:

- One loan had a property valuation grade of "C" due to secondary valuation outside of the

negative 10% tolerance.

All other loans received final credit, valuation, and regulatory compliance (wherever applicable)

grades of "A" or "B". For the above valuation grade "C" loan, we made upward adjustments to the

estimated loss coverage of 1.30x which resulted in a pool level due diligence adjustment factor of

1.00x.

R&Ws

According to our criteria (see Appendix IV of "Methodology And Assumptions For Rating U.S. RMBS

Issued 2009 And Later," published Feb. 22, 2018), S&P Global Ratings reviewed the R&Ws made by

the mortgage loan aggregator and the sponsor in this transaction. In addition, our review of the

R&W framework accounts for review triggers, knowledge qualifiers, sunset provisions, gap reps,

and enforcement mechanisms. We evaluated the strength of these R&Ws and considered whether

any breach could have a materially adverse impact on the interests of the transaction's

noteholders. If the R&Ws in the transaction documents do not address the issues in our published

R&W framework, we will determine whether we believe it is appropriate to assess additional credit

enhancement. Lastly, we will consider the R&W providers' ability to fulfill their obligations in the

event of a breach.

The seller acquired majority of the loans in this transaction from various originators and the

aggregator (Deephaven) originated or acquired the rest of the loans. The mortgage loan aggregator

does not pledge the originators' R&Ws to the trust but rather makes R&Ws on the mortgage loans

itself. For certain R&Ws that the aggregator provides as of the date it sold its loans to the sponsor

and subsequently transferred to the seller, the sponsor provides such R&Ws covering the gap

period from such sale date to the closing date (when the loans are sold to the issuing trust).

www.spglobal.com January 24, 2022 19

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

We consider the R&W framework to be weaker than those seen in recent prime jumbo transactions

because the testing of any breaches (other than any loans showing losses related to ATR) is not

automatic, but rather at the option of the controlling holder (initially, the sponsor or an affiliate of

the sponsor). The controlling noteholder's expense for the R&W review will be reimbursed from the

trust's assets. However, 25.0% or more of the noteholders will be able to initiate reviews at their

expense if the controlling holder chooses not to review or if the noteholders disagree with the

findings of the controlling holder on any review.

The R&Ws are generally consistent with our published criteria and shall remain in effect for the

transaction's life. In addition, the applicable remedy provider is required to appropriately remedy

any ensuing R&W breach if it has a materially adverse impact on the loan by either curing the

breach or purchasing the mortgage loan at the repurchase price.

The issuer will include an EPD covenant with a timeframe longer than typically seen in recent

prime jumbo transactions. An EPD event is an event that occurs with respect to a mortgage loan

when a payment (other than any payment during a forbearance period due to the COVID-19

pandemic) that is due during one of the first three months after the origination date becomes 60 or

more days' Mortgage Bankers Assn. (MBA) delinquent. An EPD mortgage loan is a mortgage loan

that has experienced an EPD event, except to the extent no scheduled monthly payment (other

than any payment during a forbearance period due to the COVID-19 pandemic) with respect to the

mortgage loan is more than 30 days' MBA delinquent on any date between the EPD event and the

sixth due date following the origination date. Within 90 days after the sponsor receives an EPD

notice, the sponsor must repurchase the related EPD mortgage loan from the issuer at a price

equal to the repurchase price.

The enforcement mechanism for R&W breaches includes provisions for a breach review, at the

controlling holder's option, by an independent reviewer or by the controlling holder itself for any

loan that experiences a realized loss. A review is mandatory only in the case of an ATR-related

realized loss. TRID defects, as determined by a judicial proceeding, will be repurchased without

any review or consideration of materiality. Dispute resolutions are ultimately subject to binding

arbitration proceedings, if necessary, to determine if a breach occurred. If the controlling holder

prevails in arbitration, then the arbitration expenses are reimbursed as part of extraordinary trust

expenses; otherwise, the expenses are not reimbursed by the trust.

Although the MOA's result reflects a solid aggregation platform, in our opinion, a party with

potentially limited repurchasing ability is providing R&Ws. Therefore, we applied a 1.10x loss

coverage adjustment to compensate for the risk associated with the financial capacities of the

R&W provider as well as the weaknesses of the framework as described above. We believe this

adjustment is appropriate in the context of the due diligence performed on the loans and the

collateral's relative credit quality.

Structural Features

Similar to some nonprime RMBS transactions issued shortly after the start of the COVID

pandemic, DRMT 2022-1 has a sequential payment structure. Principal is paid sequentially to the

senior, mezzanine, and subordinate classes.

The paying agent will make monthly interest distributions from the interest remittances and

principal from the principal remittances (see tables 7, 8, and 9).

The interest remittance amount includes the interest collected from borrowers or advanced on

their behalf (including interest payments that accompany prepayments, any compensating

interest, and interest portions of liquidation proceeds [net of expenses], subsequent recoveries,

www.spglobal.com January 24, 2022 20

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

redemption prices, and repurchase amounts) minus aggregate servicing fees, indenture trustee

fee, master servicing fees, owner trustee fees, custodial fee, the servicer advance

reimbursements, reimbursable expenses incurred by the controlling holder, and extraordinary

expenses, which are generally subject to a $500,000 annual cap. Although the extraordinary

expenses are passed through as reduced contractual interest due to the noteholders, we ran

these expenses at a certain percentage of their capped amounts to stress the excess spread (as

described in the Interest Stresses section below). We also considered the extraordinary expenses

when analyzing projected interest reduction amounts (as described in the Imputed Promises

section below).

Principal remittance amounts include the principal collected from borrowers or advanced on their

behalf (including prepayments, principal portions of liquidation proceeds [net of expenses],

subsequent recoveries, redemption prices, and repurchase amounts) minus fees, reimbursable

advances and extraordinary expenses that could not be paid from interest collections.

Table 7

Interest Payment Waterfall

Priority Payment

1 Interest and interest carry-forward amounts(i) sequentially to the class A-1, A-2, A-3, M-1, B-1,

B-2, and B-3 notes.

2 Any remaining amounts paid as part of monthly excess cash flows.

(i)Interest carry-forward amounts are deferred interest payments that accrue interest at the lower of the respective fixed coupon and the net

WAC rate. Our preliminary ratings address the full payment of all interest and interest carry-forward amounts by the final maturity date.

WAC--Weighted average coupon.

Table 8

Principal Payment Waterfall

Priority Payment

1 Unpaid interest and interest carryforward amounts to the class A-1 notes.

2 Unpaid interest and interest carryforward amounts to the class A-2 notes.

3 Principal to the class A-1 notes until reduced to zero.

4 Principal to the class A-2 notes until reduced to zero.

5 Unpaid interest and interest carryforward amounts to the class A-3 notes.

6 Principal to the class A-3 notes until reduced to zero.

7 Unpaid interest and interest carryforward amounts to the class M-1 notes.

8 Principal to the class M-1 notes until reduced to zero.

9 Unpaid interest and interest carryforward amounts to the class B-1 notes.

10 Principal to the class B-1 notes until reduced to zero.

11 Unpaid interest and interest carryforward amounts to the class B-2 notes.

12 Principal to the class B-2 notes until reduced to zero.

13 Unpaid interest and interest carryforward amounts to the class B-3 notes.

14 Principal to the class B-3 securities until reduced to zero.

15 Any remaining amounts paid as part of monthly excess cash flows.

www.spglobal.com January 24, 2022 21

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Table 9

Monthly Excess Cash Flow Waterfall

Priority Payment

1 Sequentially to class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 notes, up to the amount of any realized losses in

the current period.

2 Sequentially to class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 notes, up to the amount of any cumulative

applied realized losses until their respective note amount is reduced to zero.

3 Sequentially to class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 notes, up to the amount of any remaining

cumulative applied realized loss amounts, to reimburse for applied realized loss amounts previously

allocated thereto.

4 To the cap carry-over reserve account, up to the aggregate cap carry-over amount for class A-1, A-2, A-3,

and M-1 notes and then, sequentially, from amounts on deposit in the cap carry-over reserve account, to

class A-1, A-2, A-3, and M-1 notes any unpaid cap carry-over amounts(i).

5 To the class XS notes.

6 To the transaction parties, on a pro rata basis, any fees, expenses, or indemnification amounts not

previously paid due to application of the annual cap and any subcaps.

7 Any remaining amounts to the class R certificates.

(i)The cap carry-over amount is the positive difference between the interest that would have accrued at the fixed coupon (without regard to the

net WAC rate) and what was actually due based on the net WAC rate. Any prior unpaid cap carry-over amounts also accrue at the fixed rate. Our

preliminary ratings do not address the payment of cap carry-over amounts. WAC--Weighted average coupon.

Interest on the class A-1, A-2, A-3, and M-1 notes is based on the lower of the coupon on the notes

and the net weighted average coupon (WAC) rate (defined as the mortgage interest rate net of fees

and extraordinary expenses). In line with our imputed promises analysis, our preliminary ratings

address the lower of these two rates (see "S&P Global Ratings Definitions," published Nov. 10,

2021). Interest on class B-1, B-2 and B-3 is equal to the net WAC rate.

Under the transaction documents, the issuer can defer interest payments on these securities. A

failure to pay the interest amounts due on the securities will result in the interest being deferred.

Deferred interest (interest carry-forward amounts) accrues at the lower of the fixed rate and net

WAC rate for the class A-1, A-2, A-3, and M-1 notes, and at the net WAC rate for the class B-1, B-2,

and B-3 notes. Our preliminary ratings address ultimate P&I payments (including interest

carry-forward amounts) by the notes' final maturity date.

However, our preliminary ratings do not address the payment of cap carry-over amounts (i.e., the

difference between the fixed coupon and the net WAC rate where the fixed coupon exceeds the net

WAC rate), which are subordinated in the payment priority. In our view, neither the notes' initial

coupons nor the initial net WAC rate are de minimis, and nonpayment of the cap carryover

amounts is not considered an event of default under the transaction documents. Therefore, in line

with our analysis for imputed promises, we did not consider whether these cap carry-over

amounts are paid in our cash flow analysis.

All of the notes are paid principal sequentially. Unlike the credit enhancement seen in

shifting-interest RMBS structures, which may deplete due to scheduled and unscheduled

principal paid to the subordinate classes, the credit enhancement in DRMT 2022-1 does not

deplete since no principal payments are made on a class while a more senior class is outstanding.

The transaction starts with a 29.20% enhancement for the senior-most class and with an 17.00%

enhancement for the senior classes (rated 'AAA' through 'A') as a group. This enhancement grows

as a percentage of the current balance as the most senior class is paid down.

If the notes' aggregate class balance exceeds the pool balance, the resulting excess (the applied

www.spglobal.com January 24, 2022 22

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

realized loss amount) is applied reverse-sequentially to the class B-3, B-2, B-1, M-1, A-3, A-2, and

A-1 notes until each class' principal balance has been reduced to zero.

Additionally, if the pool balance exceeds the notes' aggregate class balance, the resulting excess

(the note write-up amount) will be applied sequentially to the class A-1, A-2, A-3, M-1, B-1, B-2,

and B-3 notes to write-up any classes (that had earlier been written-down) up to the realized and

applied realized loss amount allocated to that class.

Cash Flow And Scenario Analysis

We reviewed the transaction structure and performed a cash flow analysis to simulate various

rating stress scenarios to determine the preliminary ratings for each class consistent with our

criteria, accounting for the available credit enhancement (see table 10). We analyzed a variety of

scenarios for each rating category, including combinations of:

- Front and back-loaded default timing curves;

- Two-year recovery lag assumptions;

- Fast and slow prepayment assumptions;

- High, low, and forward/flat interest rate curve assumptions;

- WAC deterioration stress; and

- Extraordinary trust expense stresses.

Table 10

Cash Flow Assumptions

Scenario

AAA AA A BBB BB B-

Recovery lag (mos.) 24 24 24 24 24 24

Prepayments (%)(i)

Low CPR 1 2 3 4 5 6

High CPR 20 20 20 20 20 20

Extraordinary trust expenses (% of capped amount)(ii) 100.00 100.00 95.00 40.00 30.00 17.50

Foreclosure frequency (%) 53.78 47.74 38.64 29.89 20.98 9.94

Loss severity (%) 56.15 51.01 41.28 35.46 30.74 26.66

Loss coverage (%) 30.20 24.35 15.95 10.60 6.45 2.65

(i)Using a standard prepayment convention. CPR--Conditional prepayment rate. (ii)Applied monthly between periods 13 and 60 period.

CPR--Conditional prepayment rate.

Notwithstanding the use of excess interest within the transaction structure, we applied front- and

back-loaded rather than bulleted (e.g., semiannual or annual lump sum) default timing curves in

our analysis. This reflects our view of the potential volatility of cash flows given the loans are

generally recently originated or acquired by a reviewed aggregator, subject to third-party due

diligence, include structural considerations, and partial P&I advancing by the servicer.

We applied the foreclosure frequencies, loss severities, and combinations of the stresses noted

above in our cash flow runs, and observed some periodic missed interest due to the liquidity stress

associated with no advancing. To pass our rating category specific stresses, the interest deferrals

www.spglobal.com January 24, 2022 23

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

(or interest carry-forward amounts) resulting from any missed interest payments on the securities

have to be paid in full by the maturity date. All deferred interest was paid back with interest under

the applicable rating-specific stresses in our cash flow projections. The results show that each

preliminary rated class in the transaction is enhanced to a degree consistent with the respective

assigned preliminary rating.

Table 11

Structural Assessment

Class Rating

Initial class

size (%)

Initial credit

enhancement (%)

Loss coverage

(%)

Percentage point difference between credit

enhancement and loss coverage

A-1 AAA (sf) 70.80 29.20 30.20 (1.00)

A-2 AA (sf) 4.80 24.40 24.35 0.05

A-3 A (sf) 7.40 17.00 15.95 1.05

M-1 BBB

(sf)

5.60 11.40 10.60 0.80

B-1 BB (sf) 4.40 7.00 6.45 0.55

B-2 B- (sf) 4.35 2.65 2.65 0.00

B-3 NR 2.65 0.00 N/A N/A

NR--Not rated. N/A--Not applicable.

Servicer Stop Advance Stresses

Although the transaction documents provide for up to six months of P&I advance obligation, we

assumed that no P&I advances were being made in our cash flow projections on defaulted loans

that have not yet been liquidated (we assume a 24-month lag between default and liquidation).

Our cash flow projections consider this additional liquidity stress and the transaction's ability to

make monthly interest payments and, if necessary, deferred interest payments with interest

thereon (interest carryforward amounts) by the final payment date on the preliminary rated

classes.

WAC deterioration stress

To address the potential for a pool's WAC to decline over time as higher coupon loans prepay or

default, we stress the pool's projected cash flows by reducing the interest accrued on the assets.

Where appropriate, we review the distribution of loan coupons in the pool, based on measures

such as the standard deviation, interquartile range, and maximum/minimum ranges, to assess the

pool's homogeneity with respect to loan coupons.

Generally, the stress is based on the pool's WAC at the time of analysis versus 10 years later,

based on an assumed reduction in the pool balance of 10.00% per year applied to the loans with

the highest coupons. This WAC difference is the maximum WAC deterioration assumed for the

pool. The stress applied starts at zero in the transaction's first month and increases linearly each

month to the maximum through year 10, at which point, it remains constant at the maximum

through the deal's remaining life. This stress is applied in all cash flow stress scenarios at all

rating levels. For this mortgage pool, we applied a maximum WAC deterioration of 0.96%.

www.spglobal.com January 24, 2022 24

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Interest stresses

In this transaction, extraordinary trust expense payments reduce the net WAC rate, which

effectively allocates the extraordinary trust expenses pro rata across all senior and subordinate

noteholders by reducing their interest payments by the amount of the extraordinary trust

expenses paid (subject to the annual cap). Although the extraordinary expenses are passed

through as reduced contractual interest due to noteholders, we ran these expenses from periods

13 to 60 (four years) at a certain percentage of the capped amounts as specified in our criteria to

test any impact on the securities. We tested this because the securities depend on excess spread

as a form of credit enhancement and the presence of certain structural features (such as limited

P&I advancing), and because interest payments on the securities are deferrable.

Imputed Promises Analysis

We impute the interest owed to the security holders when rating U.S. RMBS transactions where

credit-related events can reduce interest owed to the tranches across the capital structure rather

than an allocation of that credit-related loss to the available credit support, based on our

guidance, see "Guidance: Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And

Later," published Dec. 8, 2020. WAC deterioration that occurs because of defaults, repurchases, or

prepayments is either already accounted for in our loss expectations or not considered

credit-related and, therefore, it is not considered as part of this analysis.

Because this transaction provides for credit-related loan modifications and extraordinary trust

expenses to reduce the net WAC at which the transaction's bond coupons are capped, we applied

the approach outlined in the guidance to assess the maximum potential rating (MPR) that could

apply based on our projected interest reduction amount (PIRA). As this is a new issue transaction,

there were no outstanding cumulative interest reduction amounts to be considered.

Consistent with our guidance, we assumed that 50.00% of the nonmodified loans projected to

default under the applicable rating stress scenario would be modified. We also assumed that 75%

of the projected modifications are interest rate modifications, with an interest rate reduction of

2%. When added to the extraordinary trust expenses, this resulted in a maximum PIRA on the

preliminary rated notes that is significantly below the 4.50% threshold. We stressed extraordinary

trust expenses by the relevant extraordinary expense application factor over 48 months, starting

from month 13 through 60 of the transaction's life. Based on the results of our analysis, there was

no impact on the securities' MPR.

Historically, we have observed that extraordinary trust expenses have been both minimal when

they occur and extremely limited in pre-2009 RMBS transactions. We continue to expect their

actual occurrence in 2009 and later transactions to be rare.

Operational Risk Assessment

Our criteria "Global Framework For Assessing Operational Risk In Structured Finance

Transactions," published Oct. 9, 2014, present our methodology and assumptions for assessing

certain operational risks (severity, portability, and disruption risks) associated with asset types

and key transaction parties (KTPs) that provide an essential service to a structured finance issuer.

According to the criteria, we cap the ratings on a transaction if we believe operational risk could

lead to credit instability and affect the ratings.

www.spglobal.com January 24, 2022 25

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

As provided in the operational risk criteria, for severity risk and portability risk, there are three

possible rankings: high, moderate, or low. For disruption risk, there are four possible rankings:

very high, high, moderate, or low. The rankings for each of the three risks determine the MPR that

can be assigned to a structured finance security for a given KTP before giving consideration to any

provisions for a backup KTP, such as a master servicer.

According to our criteria, we rank severity and portability risk for nonprime residential mortgage

collateral as moderate and low, respectively. For this transaction, Computershare, as master

servicer, is the KTP.

For our disruption risk assessment of Computershare, we determined that the company generally

exhibits the characteristics that would characterize it as stable regarding its operating condition.

This part of the assessment considered S&P Global Ratings' 'BBB' rating on Computershare Ltd.

Computershare Ltd., through its Computershare Trust Co. N.A. subsidiary, has completed its

acquisition of the Wells Fargo Corporate Trust Services (WFCTS) business to strengthen its North

American corporate trust services franchise. For key performance attributes, we considered

Computershare's track record, experience in the servicing space, portfolio growth, and

transparency, as well as any regulatory or legal issues, and determined that it met the attributes

that would categorize it as satisfactory. Based on the disruption risk assessment matrix, an

operating condition of stable, combined with an assessment of key performance attributes of

satisfactory, the disruption risk for Computershare would be low. Given these risk assessments

for Computershare, our criteria do not cap the ratings on the transaction.

Related Criteria

- General Criteria: Environmental, Social, And Governance Principles In Credit Ratings, Oct. 10,

2021

- Criteria | Structured Finance | General: Global Framework For Payment Structure And Cash

Flow Analysis Of Structured Finance Securities, Dec. 22, 2020

- Criteria | Structured Finance | General: Methodology To Derive Stressed Interest Rates In

Structured Finance, Oct. 18, 2019

- Criteria | Structured Finance | Legal: U.S. Structured Finance Asset Isolation And

Special-Purpose Entity Criteria, May 15, 2019

- Criteria | Structured Finance | General: Counterparty Risk Framework: Methodology And

Assumptions, March 8, 2019

- Criteria | Structured Finance | General: Incorporating Sovereign Risk In Rating Structured

Finance Securities: Methodology And Assumptions, Jan. 30, 2019

- Criteria | Structured Finance | RMBS: Assumptions Supplement For Methodology And

Assumptions For Rating U.S. RMBS Issued 2009 And Later, Feb. 22, 2018

- Criteria | Structured Finance | RMBS: Methodology And Assumptions For Rating U.S. RMBS

Issued 2009 And Later, Feb. 22, 2018

- Criteria | Structured Finance | RMBS: U.S. Residential Mortgage Operational Assessment

Ranking Criteria, Feb. 22, 2018

- General Criteria: Methodology For Linking Long-Term And Short-Term Ratings, April 7, 2017

- Criteria | Structured Finance | General: Global Framework For Assessing Operational Risk In

Structured Finance Transactions, Oct. 9, 2014

www.spglobal.com January 24, 2022 26

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

- General Criteria: Criteria For Assigning 'CCC+', 'CCC', 'CCC-', And 'CC' Ratings, Oct. 1, 2012

- General Criteria: Principles Of Credit Ratings, Feb. 16, 2011

- Criteria | Structured Finance | General: Methodology For Servicer Risk Assessment, May 28,

2009

Related Research

- A Sudden Correction To Fast-Rising U.S. Home Prices Isn’t Likely, Dec. 2, 2021

- Economic Outlook U.S. Q1 2022: Cruising At A Lower Altitude, Nov. 29, 2021

- S&P Global Ratings Definitions, Nov. 10, 2021

- Servicer Evaluation: Selene Finance L.P., Oct. 21, 2021

- Select Servicer List, Oct. 13, 2021

- S&P Global Ratings Publishes List Of Third-Party Due Diligence Firms Reviewed For U.S. RMBS

As Of Sept. 10, 2021, Sept. 10, 2021

- Non-Qualified Mortgage Summer Snapshot, July 22, 2021

- Computershare Ltd., July 9, 2021

- ESG Industry Report Card: Residential Mortgage-Backed Securities, March 31, 2021

- S&P Global Ratings Is Assessing The Impact Of COVID-19 On Mortgage Market Outlooks For

Global RMBS, April 17, 2020

- Servicer Evaluation: Shellpoint Mortgage Servicing, April 1, 2020

- U.S. Residential Mortgage Input File Format For LEVELS, March 6, 2020

- Credit Rating Model: LEVELS Model For U.S. Residential Mortgage Loans, Aug. 5, 2019

- Key Factors For Assessing U.S. Non-Qualified Mortgage Bank Statement Loans, April 10, 2019

- Credit Rating Model: Intex RMBS Cash Flow Model, April 7, 2017

- Older RMBS Transactions Face Increased Tail Risk As Their Pools Shrink, Aug. 9, 2012

The authors would like to thank Sagar Kapadi his research contributions to this report.

www.spglobal.com January 24, 2022 27

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

Contact List

PRIMARY CREDIT ANALYST SECONDARY CONTACT SECONDARY CONTACT

Manmadh K Venkatesan, CFA

Toronto

+ 1 (212) 438 4569

[email protected]

Kalpesh S Ghule

Centennial

+ 1 (303) 721 4157

[email protected]

Diane Lebowitz

New York

+ 212-438-1524

[email protected]

SECONDARY CONTACT SECONDARY CONTACT SURVEILLANCE CREDIT ANALYST

Kimball Ng

New York

+1 212-438-2250

[email protected]

Jessica Barbara

New York

+ 1 (212) 438 1726

[email protected]

Truc T Bui

San Francisco

+ 1 (415) 371 5065

[email protected]

ANALYTICAL MANAGER

Vanessa Purwin

New York

+ 1 (212) 438 0455

[email protected]

www.spglobal.com January 24, 2022 28

on the last page.

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1

S&P may receive compensation for its ratings and certain credit-related analyses, normally from issuers or underwriters of securities or from obligors.

S&P reserves the right to disseminate its opinions and analyses. S&P's public ratings and analyses are made available on its Web sites,

www.spglobal.com/ratings (free of charge), and www.ratingsdirect.com (subscription), and may be distributed through other means, including via

S&P publications and third-party redistributors. Additional information about our ratings fees is available at www.spglobal.com/usratingsfees.

S&P keeps certain activities of its business units separate from each other in order to preserve the independence and objectivity of their respective

activities. As a result, certain business units of S&P may have information that is not available to other S&P business units. S&P has established

policies and procedures to maintain the confidentiality of certain non-public information received in connection with each analytical process.

Credit-related and other analyses, including ratings, and statements in the Content are statements of opinion as of the date they are expressed and

not statements of fact. S&P's opinions, analyses and rating acknowledgment decisions (described below) are not recommendations to purchase,

hold, or sell any securities or to make any investment decisions, and do not address the suitability of any security. S&P assumes no obligation to

update the Content following publication in any form or format. The Content should not be relied on and is not a substitute for the skill, judgment and

experience of the user, its management, employees, advisors and/or clients when making investment and other business decisions. S&P does not act

as a fiduciary or an investment advisor except where registered as such. While S&P has obtained information from sources it believes to be reliable,

S&P does not perform an audit and undertakes no duty of due diligence or independent verification of any information it receives. Rating-related

publications may be published for a variety of reasons that are not necessarily dependent on action by rating committees, including, but not limited

to, the publication of a periodic update on a credit rating and related analyses.

To the extent that regulatory authorities allow a rating agency to acknowledge in one jurisdiction a rating issued in another jurisdiction for certain

regulatory purposes, S&P reserves the right to assign, withdraw or suspend such acknowledgment at any time and in its sole discretion. S&P Parties

disclaim any duty whatsoever arising out of the assignment, withdrawal or suspension of an acknowledgment as well as any liability for any damage

alleged to have been suffered on account thereof.

No content (including ratings, credit-related analyses and data, valuations, model, software or other application or output therefrom) or any part

thereof (Content) may be modified, reverse engineered, reproduced or distributed in any form by any means, or stored in a database or retrieval

system, without the prior written permission of Standard & Poor's Financial Services LLC or its affiliates (collectively, S&P). The Content shall not be

used for any unlawful or unauthorized purposes. S&P and any third-party providers, as well as their directors, officers, shareholders, employees or

agents (collectively S&P Parties) do not guarantee the accuracy, completeness, timeliness or availability of the Content. S&P Parties are not

responsible for any errors or omissions (negligent or otherwise), regardless of the cause, for the results obtained from the use of the Content, or for

the security or maintenance of any data input by the user. The Content is provided on an “as is” basis. S&P PARTIES DISCLAIM ANY AND ALL EXPRESS

OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE OR

USE, FREEDOM FROM BUGS, SOFTWARE ERRORS OR DEFECTS, THAT THE CONTENT'S FUNCTIONING WILL BE UNINTERRUPTED OR THAT THE

CONTENT WILL OPERATE WITH ANY SOFTWARE OR HARDWARE CONFIGURATION. In no event shall S&P Parties be liable to any party for any direct,

indirect, incidental, exemplary, compensatory, punitive, special or consequential damages, costs, expenses, legal fees, or losses (including, without

limitation, lost income or lost profits and opportunity costs or losses caused by negligence) in connection with any use of the Content even if advised

of the possibility of such damages.

Standard & Poor’s | Research | January 24, 2022 29

2784494

Presale: Deephaven Residential Mortgage Trust 2022-1