Ally Auto Receivables Trust

Presale: Ally Auto Receivables Trust 2017-4 This presale report is based on information as of Aug. 10, 2017. 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. Preliminary Ratings Class Preliminary rating(i) Type Interest rate(ii) Preliminary amount (mil. $) A-1 A-1+ (sf) Senior Fixed 280.00 A-2 AAA (sf) Senior Fixed 337.00 A-3 AAA (sf) Senior Fixed 337.00 A-4 AAA (sf) Senior Fixed 91.83 B AA+ (sf) Subordinate Fixed 23.27 C AA (sf) Subordinate Fixed 19.39 D A+ (sf) Subordinate Fixed 14.40 (i)the rating on each class of securities is preliminary and subject to change at any time. (ii)the actual coupons of the tranches will be determined on the pricing date. Profile Expected closing date Aug. 23, 2017. Collateral Sponsor and acquirer Depositor and AART certificateholder Prime auto loan receivables. Ally Bank. Ally Auto Assets LLC. Issuing entity Ally Auto Receivables Trust 2017-4. Servicer Subservicer Indenture trustee Owner trustee Ally Bank. Ally Servicing LLC. Deutsche Bank Trust Co. Americas. (A-/Negative/A-2). BNY Mellon Trust of Delaware. (AA-/Stable/A-1+). Primary Credit Analyst: Rahel Avigdor, New York (1) 212-438-4067; rahel.avigdor@spglobal.com See complete contact list on last page(s) WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 1

Profile (cont.) Lead underwriters Citigroup Global Markets Inc., Deutsche Bank Securities Inc., and SG Americas Securities LLC. AART--Ally Auto Receivables Trust. Credit Enhancement Summary AART 2017-4 AART 2017-3 Class A credit enhancement ('AAA (sf)') Initial(i) Target(i) Floor(i) Initial(i) Target(i) Floor(i) Subordination (%) 5.15 5.15 5.15 5.15 5.15 5.15 Overcollateralization (%) 0.45 1.30 1.30 0.45 1.30 1.30 Reserve account (%) 0.25 0.25 0.25 0.25 0.25 0.25 Total (%) 5.85 6.70 6.70 5.85 6.70 6.70 Class B credit enhancement ('AA+ (sf)')) Subordination (%) 3.05 3.05 3.05 3.05 3.05 3.05 Overcollateralization (%) 0.45 1.30 1.30 0.45 1.30 1.30 Reserve account (%) 0.25 0.25 0.25 0.25 0.25 0.25 Total (%) 3.75 4.60 4.60 3.75 4.60 4.60 Class C credit enhancement ('AA (sf)') Subordination (%) 1.30 1.30 1.30 1.30 1.30 1.30 Overcollateralization (%) 0.45 1.30 1.30 0.45 1.30 1.30 Reserve account (%) 0.25 0.25 0.25 0.25 0.25 0.25 Total (%) 2.00 2.85 2.85 2.00 2.85 2.85 Class D credit enhancement ('A+ (sf)') Subordination (%) 0.00 0.00 0.00 0.00 0.00 0.00 Overcollateralization (%) 0.45 1.30 1.30 0.45 1.30 1.30 Reserve account (%) 0.25 0.25 0.25 0.25 0.25 0.25 Total (%) 0.70 1.55 1.55 0.70 1.55 1.55 Estimated annual excess spread (%)(ii) 2.68 2.38 Initial aggregate gross principal balance ($) 1,107,866,077 1,057,998,088 Initial overcollateralization ($) 4,976,077 4,768,088 Total securities issued ($) 1,102,890,000 1,053,230,000 (i)the percentage of the initial aggregate principal balance. (ii)assumes a 1.00% annual servicing fee. AART--Ally Auto Receivables Trust. N/A--Not applicable. Rationale The preliminary ratings assigned to Ally Auto Receivables Trust 2017-4's (AART 2017-4's) $1.103 billion asset-backed notes series 2017-4 reflect: The availability of approximately 8.0%, 6.4%, 5.1%, and 4.1% credit support for the class A, B, C, and D notes, respectively, based on stressed cash flow scenarios. These credit support levels provide more than 5.00x, 4.50x, 4.00x, and 3.33x our 0.95%-1.05% expected net loss range to the class A, B, C, and D notes, respectively (see the WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 2

Cash Flow Modeling Assumptions And Results section for details). Our expectation of timely interest and principal payments under stressed cash flow modeling scenarios that are appropriate for the assigned preliminary ratings. Our expectation that under a moderate ('BBB') stress scenario, all else being equal, our ratings on the class A and B notes ('AAA (sf)' and 'AA+ (sf)', respectively) will remain at the assigned preliminary ratings and our ratings on the class C and D notes ('AA (sf)' and 'A+ (sf)', respectively) will remain within one rating category of the assigned preliminary ratings. These rating movements are within the limits specified by our credit stability criteria (see "Methodology: Credit Stability Criteria," published May 3, 2010). Credit enhancement in the form of subordination, overcollateralization, a reserve account, and excess spread (see the Credit Enhancement Summary table above for details). The collateral pool's characteristics. The transaction's payment and legal structures. Transaction Overview AART 2017-4 is Ally Bank's 33rd prime retail auto transaction issued under its AART term securitization program and the fourth in 2017. Ally Bank issued its first prime retail auto securitization, AART 2009-A, in September 2009. Similar to Ally Bank's previous transactions, AART 2017-4's collateral pool will consist of prime fixed-rate retail vehicle installment sales contracts originated by Ally Bank that are secured by new and used vehicles. Consistent with previous AART transactions, the minimum obligor FICO score is 660. The AART 2017-4 transaction is structured as a true sale of the receivables from Ally Bank (the sponsor and acquirer) to Ally Auto Assets LLC (the depositor and a bankruptcy-remote special-purpose entity). Ally Auto Assets, in turn, will sell the receivables as a true sale to AART 2017-4, the issuing entity. AART 2017-4 will issue four classes of notes totaling $1.103 billion (see chart 1 for the transaction structure). WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 3

Interest payments will be made on the 15th of each month (or the next business day) beginning Sept. 15, 2017. In rating this transaction, S&P Global Ratings will review the legal matters that it believes are relevant to its analysis, as outlined in its criteria. Changes From The AART 2017-3 Transaction There are no structural or credit enhancement changes from the AART 2017-3 transaction. The collateral composition changes from the AART 2017-3 transaction are: The weighted average APR increased to 5.61% from 5.35%. The weighted average seasoning decreased to 13.30 from 14.49 months. The weighted average FICO fell slightly to 737 from 738. The percentage of subvened loans decreased to 3.62% from 4.89%. WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 4

The weighted average loan-to-value (LTV) decreased to 93.29% from 93.94%. The percentage of loans with an original term of 61-75 months decreased to 65.52% from 66.06% (61-72: decreased to 53.85% from 55.06%, while 73-75 months terms increased to 11.67% from 11.00%). Loans with a remaining term of 61-72 months increased to 34.27% from 33.26% (loans with remaining term 73-75 months stable at 0.33% compared to 0.38%). We view the collateral characteristics as comparable to the series 2017-3 pool, and as a result, we have maintained our expected cumulative net loss (CNL) range at 0.95%-1.05% for the series 2017-4 pool. See the S&P Global Ratings' Expected Loss section. Transaction Structure AART 2017-4 incorporates the following structural features: A sequential-pay mechanism that will result in increased credit enhancement for the senior notes as the pool amortizes. A nonamortizing reserve account that will equal 0.25% of the initial collateral balance. Nonamortizing overcollateralization that will build to a target level by using any excess spread available after covering net losses to pay principal on the outstanding notes. Similar to series 2017-3, AART 2017-4 does not incorporate a yield supplement overcollateralization amount, given the reduction in the percentage of subvened loans. In addition to subordination, the target overcollateralization amount will equal 1.30% of the initial receivables balance. The overcollateralization amount will begin at 0.45% of the initial collateral balance and build to the 1.30% target level by using any excess spread available after covering net losses to pay principal on the notes. The required overcollateralization amount will be 1.30% of the initial receivables balance until the rated notes are paid in full. As a result, overcollateralization will grow as a percentage of the outstanding receivables as the pool amortizes. Payment Structure The payment priority that Ally Bank, the servicer, presented to us provides that the indenture trustee will use the auto loan collections to make distributions, except after the notes' acceleration following an event of default, in a specified priority (see table 1). In addition, the funds in the reserve account will be available to cover the servicing fee and the payments that occur before deposits into the reserve account. Table 1 Payment Waterfall Priority Payment 1 The 1.00% annual servicing fee. 2 Fees to the asset representations reviewers, capped at $275,000 per year. 3 Class A note interest. 4 The first-priority principal payment (if the class A notes' aggregate note principal balance is greater than the aggregate principal balance). WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 5

Table 1 Payment Waterfall (cont.) Priority Payment 5 Class B note interest. 6 The second-priority principal payment (if the class A and B notes' aggregate note principal balance is greater than the aggregate principal balance after making any first-priority principal payments). 7 Class C note interest. 8 The third-priority principal payment (if the class A, B, and C notes' aggregate principal balance is greater than the aggregate principal balance after making any first- and second-priority principal payments). 9 Class D note interest. 10 The fourth-priority principal payment (if the class A, B, C, and D notes' aggregate principal balance is greater than the aggregate principal balance after making any first-, second-, and third-priority principal payments). 11 Deposit funds into the reserve account until the balance equals the specified reserve account balance. 12 Principal on the notes(i) in an amount equal to the lesser of the following: the notes' aggregate principal balance minus the principal amounts allocated to the notes in items 3, 5, 7, and 9 above and/or the excess of the notes' aggregate principal balance; minus an amount equal to the aggregate receivables principal balance; minus the overcollateralization target amount; minus the principal amounts allocated to the notes in items 3, 5, 7, and 9 above. 13 To the indenture trustee, an amount to repay any costs the indenture trustee incurred regarding the servicer's resignation and a successor servicer's appointment. 14 To the indenture trustee, owner trustee, the administrator, and the asset representations reviewer, amounts owed under the indenture, trust agreement, and administration agreement that have not been previously paid. 15 Any remaining amounts to the certificateholders. (i)on each distribution date, the trust will apply the amounts available to make principal payments on the notes sequentially to the class A, then B, then C, and then D notes until each class is paid in full. If the AART 2017-4 pool's aggregate note balance exceeds the aggregate pool balance, the trust will use the principal collections and the excess spread to pay down the notes until the aggregate note balance equals the aggregate pool balance (items 4, 6, 8, and 10 in the payment waterfall). The amounts paid in item 11 in the payment waterfall are designed to use all principal collections and available excess spread to reduce the outstanding principal balance to its targeted level. Managed Portfolio As of June 30, 2017, Ally Bank had 2.213 million retail contracts outstanding--an approximately 2.92% increase from the same time last year (see table 2). For the six months ended June 30, 2017, the portfolio's net losses as a percentage of the average net receivables were 0.73%, higher than the 0.42% net losses as the year before. For the same time period, the portfolio's 30-plus-day delinquencies as a percentage of the month-end receivables balance were 1.51%, higher than 1.04% from the year before. Per Ally Bank, increases in delinquencies and losses beginning in 2011 and continuing through 2017 are consistent with their expectations and reflect a growing asset base and more balanced and profitable asset composition. Both Ally Bank and Ally Servicing LLC, an Ally Financial Inc. wholly owned subsidiary formerly known as Semperian LLC, service the portfolio. While we continue to observe an increase in delinquencies and losses in the managed portfolio, we view the 2017-4 pool as a better collateral mix as evidenced by the higher percentage of new vehicles (70% compared to approximately 50% in the managed portfolio). WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 6

Table 2 Ally Bank U.S. Auto Managed Portfolio Six months ended June 30 Year ended Dec. 31 Total retail contracts outstanding at end of the period (excluding bankruptcies) (in thousands) 2017 2016 2016 2015 2014 2013 2012 2011 2010 2,213 2,150 2,147 1,973 1,784 1,654 1,524 1,227 834 New vehicles 1,114 1,241 1,171 1,204 1,171 1,174 1,134 968 1,171 Used vehicles 1,099 909 976 769 613 480 390 259 976 Delinquencies (%)(i) 31-60 days 1.22 0.85 1.26 0.89 0.80 0.69 0.58 0.44 0.27 61-90 days 0.24 0.16 0.26 0.19 0.15 0.14 0.13 0.06 0.05 90-plus days 0.05 0.03 0.06 0.03 0.03 0.03 0.02 0.01 0.01 Total delinquencies 1.51 1.04 1.58 1.11 0.98 0.85 0.73 0.51 0.33 Repossessions as a % of the avg. no. of contracts outstanding (including bankruptcies) 1.54 1.01 1.15 0.84 0.77 0.66 0.51 0.45 0.37 Net losses as a % of the avg. net receivables 0.73 0.42 0.54 0.34 0.31 0.31 0.23 0.19 0.17 (i)the month-end delinquencies in dollars as a percentage of the month-end receivables balance. Collateral Pool Analysis We analyzed the AART 2017-4 collateral pool with those of Ally Bank's previous securitization pools (see tables 3 and 4 for AART retail auto securitizations since 2012). The AART securitized pools have an approximately 737-761 FICO range, an approximately 92%-98% LTV range, and an approximately 67%-92% new vehicle composition range. Although the AART pools' weighted average FICO has generally decreased since the series 2012-1 transaction, it has remained generally stable since the series 2015-2 transaction. The weighted average LTV generally increased between 2012 and the beginning of 2015, and then started decreasing. Table 3 Collateral Comparison(i) AART 2017-4 2017-3 2017-2(ii) 2017-1(ii) 2016-3 2016-2 2016-1(ii) 2015-2 2015-1 Pool size (mil. $) 1,107.9 1,058.0 1,019.9 1,116.6 516.2 1,034.8 1,066.7 1,094.6 1,035.9 No. of receivables 70,003 67,797 67,771 69,007 37,913 57,588 58,961 59,975 52,991 Avg. principal balance ($) Weighted avg. APR excluding the YSOA (%) Weighted avg. original term (mos.) Weighted avg. remaining term (mos.) 15,826 15,605 15,049 16,181 13,615 17,970 18,092 18,251 19,549 5.61 5.35 5.39 5.26 5.12 4.36 4.27 4.34 3.80 65.59 65.93 65.45 66.01 65.31 66.12 65.55 65.95 66.19 52.30 51.45 52.20 52.81 50.90 54.15 53.95 52.31 52.96 Seasoning (mos.) 13.30 14.49 13.25 13.20 14.41 11.97 11.60 13.61 13.23 WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 7

Table 3 Collateral Comparison(i) (cont.) AART Total % of loans with an original term of 60-plus mos. Total % of loans with an original term of 73-75 mos. 2017-4 2017-3 2017-2(ii) 2017-1(ii) 2016-3 2016-2 2016-1(ii) 2015-2 2015-1 65.52 66.06 65.46 68.12 68.70 68.03 63.91 65.62 66.35 11.67 11.00 10.71 9.60 9.96 9.95 9.78 9.94 9.63 New vehicles (%) 70.00 70.00 70.00 70.08 70.11 70.01 70.73 70.90 70.69 Used vehicles (%) 30.00 30.00 30.00 29.92 29.89 29.99 29.27 29.10 29.31 Subvened loans (%) (iii) Nonsubvened loans (%) Weighted avg. original FICO score 3.62 4.89 5.49 6.38 9.20 20.03 20.15 18.26 20.48 96.38 95.11 94.51 93.62 90.80 79.97 79.85 81.74 79.52 737.04 738.00 739.00 740.16 739.74 740.05 740.71 740.2 746.0 Weighted avg. LTV 93.29 93.94 93.30 94.40 92.20 94.65 94.35 95.79 95.08 GM-related vehicles (%) Chrysler-related vehicles (%) Top five state concentrations (%) Original ECNL range (%) Revised ECNL range (as of June 2017) (%) 42.64 45.05 43.87 45.60 48.9 61.8 62.5 63.2 64.8 TX=12.69 17.3 16.8 16.0 15.3 15.0 15.2 12.9 14.6 14.4 TX= 12.66 TX= 12.45 TX= 12.85 TX=11.48 TX=12.60 TX=12.65 TX=13.59 TX=13.97 CA=9.56 CA= 9.19 CA= 9.25 CA= 9.73 CA=9.01 CA=8.49 CA=8.83 CA=7.91 CA=7.24 FL=8.35 FL= 8.52 FL= 8.77 FL= 8.32 FL=8.62 FL=7.51 FL=7.58 FL=7.12 FL=7.20 IL=4.89 PA= 5.13 PA= 5.44 PA= 5.48 PA=5.73 PA=5.44 PA=5.58 PA=5.18 PA=5.13 PA=4.84 IL= 5.06 IL= 5.01 IL= 4.99 IL=5.28 IL=5.03 IL=5.02 IL=4.74 IL=4.61 0.95-1.05 0.95-1.05 N/A N/A 0.80-0.90 0.80-0.90 N/A 0.75-0.85 0.70-0.80 N/A N/A N/A N/A N/A 0.80-0.90 N/A 0.75-0.85 0.45-0.55 (i)all percentages are of the initial gross receivables balance. (ii)not rated by S&P Global Ratings. (iii)the subvened loans are loans that Ally Bank originated or acquired under special incentive rate financing programs. AART--Ally Auto Receivables Trust. APR--Annual percentage rate. YSOA--Yield supplement overcollateralization amount. LTV--Loan-to-value. ECNL--Expected cumulative net loss. Table 4 Collateral Comparison(i) AART 2014-3 2014-2 2014-1 2014-A 2013-2(ii) 2013-1 2012-5 2012-4 2012-A Pool size (mil. $) 1,025.0 1,530.9 2,048.9 1,125.0 1,107.0 1,292.0 1,385.1 1,405.2 2,000.0 No. of receivables 45,620 70,196 95,976 51,875 49,460 58,552 56,379 57,516 84,067 Avg. principal balance ($) Weighted avg. APR excluding the YSOA (%) 22,469 21,809 21,348 21,687 22,382 22,066 24,568 24,432 23,791 3.87 4.13 3.94 3.83 3.61 3.53 3.17 3.07 3.05 WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 8

Table 4 Collateral Comparison(i) (cont.) AART Weighted avg. original term (mos.) Weighted avg. remaining term (mos.) 2014-3 2014-2 2014-1 2014-A 2013-2(ii) 2013-1 2012-5 2012-4 2012-A 66.65 66.94 65.74 65.99 65.54 64.78 64.20 63.57 63.11 57.59 58.01 54.66 54.92 55.58 54.00 58.71 58.87 55.78 Seasoning (mos.) 9.06 8.94 11.08 11.06 9.96 10.78 5.49 4.70 7.34 Total % of loans with an original term of 60-plus mos. Total % of loans with an original term of 73-75 mos. 68.69 70.31 63.57 64.95 61.97 58.93 56.18 52.95 50.27 9.60 15.42 10.01 9.87 10.16 9.47 0.00 0.00 0.00 New vehicles (%) 70.20 66.61 73.91 75.21 77.89 80.41 82.13 84.06 86.00 Used vehicles (%) 29.80 33.39 26.09 24.79 22.11 19.59 17.87 15.94 14.00 Subvened loans (%)(iii) 23.67 13.18 24.86 30.19 34.63 42.62 44.10 47.90 48.79 Nonsubvened loans (%) Weighted avg. original FICO score 76.33 86.82 75.14 69.81 65.37 57.38 55.90 52.10 51.21 741.4 749.6 745.8 744.4 742.9 744.7 749.5 751.8 760.7 Weighted avg. LTV 97.37 97.75 95.82 96.40 96.15 96.38 94.45 94.05 93.00 GM-related vehicles (%) Chrysler-related vehicles (%) Top five state concentrations (%) Original ECNL range (%) Revised ECNL range (as of June 2017) (%) 66.4 65.0 67.1 68.2 67.2 63.2 67.2 66.6 27.5 15.9 17.8 18.4 18.0 20.5 25.6 23.4 23.6 64.5 TX=14.22 TX=13.50 TX=13.92 TX=13.28 TX=13.38 TX=12.18 TX=13.86 TX=14.39 TX=12.52 CA=7.25 FL=7.57 FL=7.16 FL=7.06 FL=7.05 FL=7.10 FL=6.65 FL=6.47 CA=6.69 FL=7.15 CA=6.39 CA=6.34 CA=6.38 CA=5.92 PA=5.71 PA=5.69 PA=5.25 MI=5.74 PA=4.94 PA=6.06 PA=4.98 PA=4.87 PA=5.40 MI=5.12 MI=5.14 MI=5.08 FL=5.72 IL=4.82 IL=4.52 IL=4.56 MI=4.61 MI=4.83 CA=5.11 CA=4.31 CA=4.27 PA=4.69 0.80-0.90 0.70-0.80 0.75-0.9 0.75-0.9 N/A 0.75-0.9 0.75-0.9 0.75-0.9 0.85-1.0 0.65-0.75 0.40-0.50 0.45-0.55 0.60-0.70 N/A N/A N/A N/A N/A (i)all percentages are of the initial gross receivables balance. (ii)not rated by S&P Global Ratings. (iii)the subvened loans are loans that Ally Bank originated or acquired under special incentive rate financing programs. AART--Ally Auto Receivables Trust. APR--Annual percentage rate. YSOA--Yield supplement overcollateralization amount. LTV--Loan-to-value. ECNL--Expected cumulative net loss. Securitization Performance Ally Bank has been issuing auto loan asset-backed securities since 2009, and its securitizations have performed well. The transactions that have paid off (series 2009-A through 2013-1) experienced CNLs of 0.25%-0.50% (see chart 2 for CNL performance of the paid-off securitizations). WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 9

Chart 2 AART Performance: Surveillance Update S&P Global Ratings currently maintains ratings on nine AART trust transactions issued between 2012 and 2017. AART's outstanding transactions have exhibited consistent securitization performance. On June 8, 2017, we raised our ratings on eight classes from four AART trusts (series 2014-1, 2014-2, 2015-1, and 2016-2) and affirmed the remaining 21 ratings from seven AART transactions (see "Eight Ratings Raised, 21 Affirmed On Seven Ally Auto Receivables Trust Transactions," published June 8, 2017). At that time, we revised the losses on various transactions to 0.40%-0.90% (see table 5 and chart 3). In our opinion, the outstanding transactions remain adequately enhanced at the current ratings. Credit performance has been better than expected, and there has been significant growth in credit enhancement as a percentage of the current receivables balance. We will continue to monitor the outstanding transactions' performance to assess whether the credit enhancement remains commensurate with our ratings. WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 10

Table 5 Performance Data For Outstanding Ally Bank Transactions As Of The July 2017 Distribution Date AART transaction Month Pool factor (%) CNL (%) 60-plus-day delinquency (%) Initial projected lifetime CNL (%) Revised lifetime CNL expected as of June 2017 (%) 2014-A 38 13.95 0.58 0.29 0.75-0.90 0.60-0.70 2014-1 37 14.80 0.46 0.44 0.75-0.90 0.45-0.55 2014-2 34 22.70 0.38 0.22 0.70-0.80 0.40-0.50 2014-3 31 26.45 0.52 0.24 0.80-0.90 0.65-0.75 2015-1 24 35.25 0.29 0.21 0.70-0.80 0.45-0.55 2015-2 22 36.05 0.47 0.33 0.75-0.85 0.75-0.85 2016-2 17 48.48 0.37 0.24 0.80-0.90 0.80-0.90 2016-3 14 54.74 0.38 0.34 0.80-0.90 N/A 2017-3 2 92.40 0.01 0.10 0.95-1.05 N/A AART--Ally Auto Receivables Trust. CNL--Cumulative net loss. N/A--Not applicable. Chart 3 WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 11

S&P Global Ratings' Expected Loss: 0.95%-1.05% To derive our base-case loss for the AART 2017-4 transaction, we analyzed static pool originations performance data since 2003 on loans that Ally Bank and General Motors Acceptance Corp. (GMAC) had acquired, which had collateral characteristics similar to the securitized pool. The data we received were segmented by new- and used-vehicle mixes, credit grade, APR subvention, and original loan term. We developed expected net loss projections for each combination of these credit buckets. We relied on originations that had at least 12 months of performance data when determining our net loss projections. Our loss projection analysis focused on quarterly vintages that were originated primarily from 2008 to 2015. We then weighted these projections based on the actual concentration of the various segments in the AART 2017-4 pool. We also examined the CNLs and cumulative gross loss performance of Ally Bank's paid-off pools and its outstanding securitized pools. We used the paid-off deals to create a loss-timing curve and used that curve to project pool losses on the outstanding AART securitizations. We also looked at the cumulative recovery rates and delinquencies by vintage and considered our current and forward-looking views of the economy and auto loan sector (see the Securitization Performance and AART Performance: Surveillance Update sections for more information). We considered the strong performance of Ally Bank's AART retail loan securitizations, the collateral and credit grade mix of the series 2017-4 pool, the origination static pool loss analysis, and our forward-looking view of the economy (including lower expected recovery rates). The collateral characteristics are comparable with the series 2017-3 pool; as a result we have maintained our expected cumulative net loss (CNL) range at 0.95%-1.05%. Cash Flow Modeling Assumptions And Results We modeled the AART 2017-4 transaction to simulate 'AAA' stress scenarios (see table 6). In our modeling approach, we used a two-pool method in which the subvened loans prepay at much slower rates than the nonsubvened ones (for cash flow purposes, nonsubvened loans have APRs greater than 4%), and the subvened loans' loss contribution to the pool's total aggregate losses is less than their proportional representation in the pool. We believe that the lower-apr loans will prepay at a slower speed than the higher-apr loans. Table 6 Cash Flow Assumptions And Results (Two-Pool Approach) Class A B C D Scenario (preliminary rating) AAA (sf) AA+ (sf) AA (sf) A+ (sf) CNL timing by months outstanding (actual) (12/24/36/48) (%) Subvened CNL 27/60/83/100 27/60/83/100 27/60/83/100 27/60/83/100 Nonsubvened CNL 42/77/96/100 42/77/96/100 42/77/96/100 42/77/96/100 Aggregate CNL 39/74/94/100 39/74/94/100 39/74/94/100 39/74/94/100 Loss allocations (%) Subvened receivables 20 20 20 20 WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 12

Table 6 Cash Flow Assumptions And Results (Two-Pool Approach) (cont.) Class A B C D Nonsubvened receivables 80 80 80 80 ABS voluntary prepayments (%) Subvened(i) 0.00 0.00 0.00 0.00 Nonsubvened(i) 1.80 1.75 1.70 1.65 Recoveries (%) 50 50 50 50 Recovery lag (mos.) 4 4 4 4 Servicing fee (%) 1.00 1.00 1.00 1.00 Approximate break-even CNL levels (%)(ii) 8.0 6.4 5.1 4.0 (i)the subvened/nonsubvened cut-off annual percentage rate is 4.0%. (ii)the maximum CNLs on the pool that the transaction can withstand without triggering a payment default on the relevant class of notes. ABS--Absolute prepayment speed. CNL--Cumulative net loss. We stressed the weighted average APR on the AART 2017-4 pool by assuming zero prepayments on the low-apr loans and applying a slower loss curve to them. In a stressed scenario, liquidity risk could arise as a result of interest shortfalls that occur when the yield on the assets is lower than the note yield (especially because the notes pay principal sequentially) and lead to higher-coupon debt outstanding at the tail end of the transaction. We also allocated losses disproportionately between the subvened and nonsubvened loans. Ally Bank's and GMAC's origination static pool data indicate that nonsubvened loans tend to incur higher losses than subvened loans. In our internal cash flow runs, we used a 0.95%-1.05% expected CNL range and applied the stresses as outlined above. The break-even results show that each class in the AART 2017-4 transaction has sufficient credit enhancement to withstand a stressed net loss level that is consistent with the assigned preliminary ratings. Sensitivity Analysis In addition to running break-even cash flows, we ran a two-pool sensitivity analysis to see how the preliminary ratings on the notes could be affected by losses that are moderately higher than what we currently expect (see table 7 and chart 4). Table 7 Sensitivity Analysis Summary--Moderate Loss Scenario Cumulative net loss level (%) 2.00 Loss timing by months outstanding (12/24/36/48) (%) Subvened 27/60/83/100 Nonsubvened 42/77/96/100 Aggregate 39/74/94/100 Loss allocations (%) Subvened receivables 20 Nonsubvened receivables 80 WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 13

Table 7 Sensitivity Analysis Summary--Moderate Loss Scenario (cont.) Voluntary ABS (%)(i) Subvened 0.00 Nonsubvened 1.50 Servicing fee (%) 1.00 Recovery rate (%) 50 Recovery lag (mos.) 4 Haircut to excess spread (%) 10 Potential rating decline Class A Class B Class C Class D No rating decline expected No rating decline expected One rating category One rating category (i)the subvened/nonsubvened cutoff annual percentage rate is 4.0%. ABS--Absolute prepayment speed. Chart 4 WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 14

Moderate loss scenario: 2.00% (2.0x cumulative net loss results) Under the 2.00% moderate stress loss scenario (approximately 2.0x our expected loss level), we assumed loss curves on the subvened and nonsubvened loans that are similar to what we used in our 'AAA' stress scenario, and we allocated losses disproportionately between them. To stress excess spread, we assumed a 1.50% voluntary prepayment speed for loans with APRs higher than 4.00% and a zero voluntary prepayment speed for loans with APRs less than 4.00%. We also applied a 10.00% haircut to the remaining excess spread at any given month. Under this scenario, the transaction reaches the overcollateralization target and floor by month five. Interest is paid on a timely basis, and the class A, B, C, and D notes are repaid in full by months 47, 50, 53, and 55, respectively. While the notes are outstanding, excess cash flow is released in months five through 34 and again in months 37 through 44 and then in months 50 through 65, totaling approximately $28.76 million (2.1% of the initial collateral balance). Our sensitivity analysis indicates that in a moderate loss stress scenario,all else being equal, the ratings on the class A and B notes will remain at the assigned preliminary ratings, and our ratings on the class C and D notes will remain within one rating category of the assigned preliminary ratings. This is in line with our credit stability criteria, which state that the maximum deterioration associated with 'AAA' and 'AA' ratings within the first year of issuance is a one-category reduction and the maximum deterioration associated with 'A' ratings within the first year of issuance is a two-category reduction(for more information, see "Methodology: Credit Stability Criteria," May 3, 2010). Money Market Tranche Sizing The proposed money market tranche (class A-1) has a 12-month legal final maturity date (Sept. 17, 2018). To test whether the money market tranche can be repaid by then, we ran cash flows using assumptions that delay the principal collections during the 12-month time period. We assumed zero defaults, a 0.5% absolute prepayment speed (ABS) for the nonsubvened loans, and a 0.0% ABS for the subvened loans in our cash flow run. Based on our cash flow runs, approximately 12 months of principal collections would be sufficient to pay off the money market tranche. Legal Final Maturity We also tested the legal final maturity dates Ally Bank set for the intermediate, long-dated tranches (classes A-2, A-3, A-4, B, and C). In a zero-loss, zero-prepayment scenario, the class A-2, A-3, A-4, B, and C notes were fully amortized at least three months before their legal final maturity dates. For the longest-dated security, class D, at least six months was added to the tenor of the pool's longest receivable to accommodate extensions on the receivables. In our break-even cash flow scenarios, we confirmed that there was sufficient credit enhancement to both cover losses and repay the related notes in full by its legal final maturity date. Ally Bank Ally Bank was incorporated in 2004 and is a wholly owned, indirect subsidiary of Ally Financial Inc. (BB+/Stable/B). Ally Financial Inc., which is now an independently traded financial services company, is the former GMAC, which was WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 15

General Motors Corp.'s original captive finance company. Ally Bank was incorporated under the GMAC Automotive Bank name, which was later changed to GMAC Bank in 2006. In December 2008, GMAC Bank converted into a Utah-chartered commercial nonmember bank and changed its name to Ally Bank in April 2009. In March 2016, Ally Bank received approval from the Board of Governors of the Federal Reserve System (FRS) to become a state member bank and be regulated by the FRB through the Federal Reserve Bank of Chicago. It is subject to regulation by the FRB and the Utah Department of Financial Institutions. Ally Bank offers consumer retail contracts and consumer leases to and through dealerships and their customers. Ally Bank also provides dealer equipment facilities financing, residential mortgage financing, and wholesale and retail financing. Ally Bank has been actively originating assets since 2004. Its underwriting and acquisition strategy includes using internal credit scoring models to assess potential borrowers' creditworthiness and ability to pay. These scoring models consider each customer's credit profile and previous experience with Ally Bank and its affiliates, the terms of the proposed installment sales contract, the vehicle's asset value, and the potential borrower's amount of equity in the vehicle. Ally Bank evaluates and updates these scoring models periodically to account for changes in the perceived impact of specific inputs on applicants' creditworthiness. Both Ally Bank and Ally Servicing LLC service the receivables that Ally Bank originates. Ally Financial Inc. previously had preferred lender agreements with Chrysler and GM (beginning in 2009), but these expired in April 2013 and February 2014, respectively. Ally Financial Inc. continues to support auto financing with the two Detroit automakers but without an exclusive agreement to finance their respective vehicle sales incentive programs. On April 27, 2015, Ally Financial Inc. announced it will become the preferred lender for Mitsubishi Motors in the U.S., replacing the brand captive, Mitsubishi Motors Credit of America Inc. North America Inc. On May 1, 2015, Ally Financial Inc. was named the preferred lender for Aston Martin in the U.S. Related Criteria General Criteria: Methodology For Linking Long-Term And Short-Term Ratings, April 7, 2017 Criteria - Structured Finance - General: Ratings Above The Sovereign - Structured Finance: Methodology And Assumptions, Aug. 8, 2016 Criteria - Structured Finance - General: Methodology: Criteria For Global Structured Finance Transactions Subject To A Change In Payment Priorities Or Sale Of Collateral Upon A Nonmonetary EOD, March 2, 2015 Criteria - Structured Finance - General: Criteria Methodology Applied To Fees, Expenses, And Indemnifications, July 12, 2012 General Criteria: Global Investment Criteria For Temporary Investments In Transaction Accounts, May 31, 2012 Criteria - Structured Finance - RMBS: U.S. Interest Rate Assumptions Revised For May 2012 And Thereafter, April 30, 2012 Criteria - Structured Finance - ABS: General Methodology And Assumptions For Rating U.S. Auto Loan Securitizations, Jan. 11, 2011 WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 16

Criteria - Structured Finance - General: Standard & Poor's Revises Criteria Methodology For Servicer Risk Assessment, May 28, 2009 Legal Criteria: Legal Criteria For U.S. Structured Finance Transactions: Special-Purpose Entities, Oct. 1, 2006 Legal Criteria: Legal Criteria For U.S. Structured Finance Transactions: Appendix III: Revised UCC Article 9 Criteria, Oct. 1, 2006 Legal Criteria: Legal Criteria For U.S. Structured Finance Transactions: Criteria Related To Asset-Backed Securities, Oct. 1, 2006 Legal Criteria: Legal Criteria For U.S. Structured Finance Transactions: Securitizations By SPE Transferors And Non-Code Transferors, Oct. 1, 2006 Legal Criteria: Legal Criteria For U.S. Structured Finance Transactions: Select Issues Criteria, Oct. 1, 2006 Related Research Eight Ratings Raised, 21 Affirmed On Seven Ally Auto Receivables Trust Transactions, June 8, 2017 Global Structured Finance Scenario And Sensitivity Analysis 2016: The Effects Of the Top Five Macroeconomic Factors, Dec. 16, 2016 In addition to the criteria specific to this type of security (listed above), the following criteria articles, which are generally applicable to all ratings, may have affected this rating action: "Post-Default Ratings Methodology: When Does Standard & Poor's Raise A Rating From 'D' Or 'SD'?," March 23, 2015; "Global Framework For Assessing Operational Risk In Structured Finance Transactions," Oct. 9, 2014; "Methodology: Timeliness of Payments: Grace Periods, Guarantees, And Use of 'D' And 'SD' Ratings," Oct. 24, 2013; "Counterparty Risk Framework Methodology And Assumptions," June 25, 2013; "Criteria For Assigning 'CCC+', 'CCC', 'CCC-', And 'CC' Ratings," Oct. 1, 2012; "Methodology: Credit Stability Criteria," May 3, 2010; and "Use of CreditWatch And Outlooks," Sept. 14, 2009. The analyst would like to thank Kailash Ahirrao for his analytical contribution on this transaction and presale report. Analytical Team Primary Credit Analyst: Rahel Avigdor, New York (1) 212-438-4067; rahel.avigdor@spglobal.com WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 17

Copyright 2017 by Standard & Poor s Financial Services LLC. All rights reserved. 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. 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. 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 acknowledgement 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. 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 nonpublic information received in connection with each analytical process. S&P may receive compensation for its ratings and certain 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.standardandpoors.com (free of charge), and www.ratingsdirect.com and www.globalcreditportal.com (subscription) and www.spcapitaliq.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.standardandpoors.com/usratingsfees. STANDARD & POOR'S, S&P and RATINGSDIRECT are registered trademarks of Standard & Poor's Financial Services LLC. WWW.STANDARDANDPOORS.COM/RATINGSDIRECT AUGUST 10, 2017 18