BAC: Room to Run as Rising Net Interest Margin Levers over Declining Expenses and Share Count

Print Friendly, PDF & Email



January 20, 2014

BAC: Room to Run as Rising Net Interest Margin Levers over Declining Expenses and Share Count

  • On the last earnings call, BAC guided to a tangible ROE of 14% “as we look out over three years” based on a 1% ROA and tangible equity/assets ratio “that’s just over 7%” (7.5% in the last quarter). In practice, we expect this profitability to be achieved in 2016 consistent with a one-year price target of $22 being 1.5x our estimate of $15 for tangible book value (“TBV”)/share at end-2014; the target is 12x our 2015EPS estimate of $1.83 (versus consensus of $1.62).
  • Our estimates include $7bn of litigation expenses over the next two years. While the outcome could be meaningfully higher, we note an additional $10bn reduces TBV/share after tax-effects by 70 cents and hence our price target, at a 1.5x multiple, by just $1.
  • If headline risk creates volatility in the share price, we expect BAC to be a size-buyer particularly given proximity to forward TBV (forecast at $15/share for end-2014 and $16 for end-2015).
  • Our variance from consensus lies in the net interest margin (“NIM”) forecast. Consensus apparently discounts a core NIM (i.e. excluding the impact of trading and market-related effects such as MBS premium amortization) of 3.15% in 2015 reflecting a “grind” higher from 3.08% in 2013Q4 as, for example, BAC continues to manage down long-term debt. Rather than anchoring to today’s rate environment, we believe the proper default for the NIM forecast is the implied yield curve which has 6-month Libor (“6ML”) in mid-2015 at 1.5%. Banks have historically captured about half of an increase in short rates in their deposit franchises (and will not give this up on the asset side as the forward markets also indicate curve-steepening) and we expect BAC to do at least as well; that said, to be conservative, we model the NIM for 2015FY at 30 basis points higher than in 2013Q4.
  • Each 10 basis point increase in 2015FY net interest margin adds 8 cents/share to the EPS estimate. We note that for mid-2016, the forward markets are indicating 6ML at 2.5%.
  • A rising NIM at BAC is levered by declining non-interest expenses, with the benefit from re-engineering initiatives (“Project New BAC”) and the wind-down of the legacy asset services (“LAS”) business, and a declining share count as the payout ratio rises from the 35% in 2013 to 90% in 2015. Specifically (see Exhibit), we expect:
  • Expenses, excluding litigation, to fall by $7.5bn for 2015 versus 2013 of which the legacy asset services (“LAS”) business contributes over $5bn and “Project New BAC” re-engineering the balance even after allowing for an increase in revenue-related expenses.
  • The common stock count to fall 8% by end-2015 as stock buyback increases to $7bn in 2014 and $16bn in 2015 (with an average buyback price of $20 and $25 respectively). Given an assumed increase in dividend/share to 15 cents and 30 cents respectively, these buyback amounts generate respective payout ratios of 60% and 90% and assume BAC does not de-lever from the current tangible equity/assets ratio of just over 7%.
  • Beyond litigation expense, possible CCAR constraints on the payout ratio, and changes in the implied yield curve, a key risk to our thesis is around volatile revenue streams such as investment and mortgage banking and equity investment gains as well as trading activity. Between them, these contributed $24bn to, or over one-quarter of, 2013 revenues. While normal variation will affect our 2015 EPS estimate, we do not expect an enduring impact on valuation from anything but structural change and see this as unlikely given BAC’s competitive positioning and strong execution.
  • We expect a step-down in debit interchange fees (booked through otherwise relatively stable card income), from litigation and structural shifts in the payments industry, to be offset by consumer charges notwithstanding the public backlash against, and consequent withdrawal in November 2011 of, a $5 monthly debit-card fee for a debit card proposed by BAC.

Exhibit: BAC Summary Model

Investment Conclusion

BAC has guided to a 14% return-on-tangible-equity (“tangible ROE”) over a three-year time and, in practice, we expect this to be achieved by 2016. Given we expect tangible book value (“TBV”)/share of $15/share by end-2014 (when investors will be rolling-forward to 2016 estimates), this is consistent with a one-year target price of $22 being 1.5x TBV (see Exhibit 1).

Exhibit 1: Tangible Equity Returns Drive Bank Valuation

Margin Leverage

The upside opportunity at BAC arises as an increase in the net interest margin as the US rate environment normalizes is put over declining non-interest expenses, with the benefit from re-engineering initiatives (“Project New BAC”) and the wind-down of the legacy asset services (“LAS”) business, and a declining share count as the payout ratio rises from the 35% in 2013 to 90% in 2015.

  • Based on the current implied yield curve for the US, BAC will likely see its core net interest margin (i.e. excluding the effect of trading-related assets) expand from an average 3.1% in 2013 to 3.4% or more in 2015; each 10 basis point increase in this core net interest margin is worth 8 cents/share of 2015 EPS. Specifically, the forward markets are indicating that short rates, as measured by 6-month Libor, will increase from the present 33 basis points to near 1.5% by mid-2015, 2.5% by mid-2016, and 3.3% by mid-2017.
  • We expect expenses associated with the LAS to decline over $2bn and $1.5bn respectively in each of the next two years; this still leaves room for litigation expenses of $4.5bn and $2.5bn in 2014 and 2015. Given broader expense initiatives, overall non-interest expenses can fall from $69bn in 2013 to $60bn.

BAC will not capture the entire increase in rising short rates because it will need to re-price deposits and will have a negative mix-shift from low-interest to higher-interest accounts, but history indicates that banks in general tend to capture half of an increase in short rates in the early part of a tightening cycle. We expect BAC to do better but, for modeling purposes, are using a core margin of 3.4% in 2015 generating an EPS estimate of over $1.80 versus consensus of ~$1.60 (see Exhibit above and Appendix).

In the last rate-tightening cycle from 2003-2006, banks did not see rising net interest margins because rising short rates were accompanied by a flattening of the curve (as the slope between the 3-year swap rate and 6-month Libor declined from 1.3% to negative 30 basis points – see Exhibit 2). In contrast, on this occasion, the forward markets are indicating that this slope will increase from the current 0.6% to 1.1% by mid-2015 (albeit with flattening in 2016 and 2017). This is margin “nirvana” as rising short rates allow banks to monetize their deposit franchises and simultaneous curve-steepening improves reinvestment opportunities on the asset side as well.

Exhibit 2: Relationship between Core Net Interest Margin and Swap Curve for FDIC-Reporting US Banks

Source: SNL, SSR Analysis

Outcome rates, of course, can differ widely from the indications of the forward market, and these indications are by no means stable (see Exhibit 3). However, we believe they must be the default for modeling bank earnings; any alternative rate view, after all, represents an active bet in the bond market.

Exhibit 3: Forward Market Indications for 3-year swap rate (“3YS”) and 6-month Libor (“6ML”)

Source: Bloomberg

Capital Management

Investors in BAC have downside protection because the bank is highly capital-generative and committed to buying back stock. We acknowledge the risk of CCAR constraints but note that the bank presently has an accounting equity/assets and CT1 (“common Tier 1”) capital ratios of over 11%, and a fully-phased in Basel 3 CT1 ratio under the advanced and standard approaches of 10% and 9% respectively (versus the minimum requirement, which does not kick-in until 2019, of 8.5%); in addition, the bank meets the supplementary leverage ratio (“SLR”) requirements.

Assuming an accounting equity/assets ratio of 11% is sufficient, and BAC is not required to de-lever beyond this, the bank can buy back ~$7bn and ~$16bn of stock in 2014 and 2015 respectively allowing a near-8% reduction in the share count over the next two years even if the average buyback price in 2015 reaches $25 (see Exhibit 4). Our assumption that BAC does not de-lever further is equivalent to assuming that the payout ratio rises from 35% in 2013 to 90% in 2015; this, in turn, is consistent with assumed balance-sheet growth in 2015 of 1% and a return-on-equity of ~10%. With the discontinued real-estate portfolio largely run-off and the economy improving (consistent with assumed rate-tightening), we expect loan growth in 2015 of 5% but for the equivalent ~$50bn of additional balances to be offset by zero or negative growth in the debt securities and short-term investment portfolios (largely fed funds purchased and repos).

While it does not meaningfully affect the analysis, we note in passing that our assumption that the dividend rises to 15 cents/share in 2014 and 30 cents/share in 2015. The assumed 2015 dividend represents less than a 2% yield even on today’s share price and we do not expect BAC to be more aggressive given management’s indication that stock buyback is a priority and possibly greater regulatory resistance to the more lasting commitment of dividends versus the more easily-adjusted return-of-capital through buyback.

Exhibit 4: Stock Buyback at BAC

Source: Company Reports, SSR Estimates

Non-Interest Expense

Through the expense-initiative “Project New BAC”, BAC reduced the quarterly expenses excluding litigation and those associated with the legacy asset services (“LAS”) business by $900 million from $14.7bn in 2012Q1 to $13.8bn in 2013Q1; the firm has committed to a further reduction in the quarterly run-rate of $600mm by 2014Q1 (meaning that 2014Q1 expenses, excluding retirement-eligible costs of ~$900mm, will need to be sequentially flat from 2013Q4 which included $300mm of revenue-related expenses that offset Project New BAC savings despite the impact of incentive compensation which otherwise tends to lift first-quarter expenses) and a further $500mm by mid-2015. Given the offset to Project New BAC savings from revenue-related compensation, we take these commitments to mean that annual expenses will fall $1.5bn from 2013 to 2014 and a further $1.2bn from 2014 to 2015 (see Exhibit 5).

Exhibit 5: Stock Buyback at BAC

Other key components of non-interest expenses are those related to the LAS business and litigation expenses. BAC has guided LAS expenses to $1.1bn for 2014Q4 and $500mm for 2015Q4 and noted that litigation expense is inherently unpredictable; for definiteness, we assume a decline of $1.5bn in 2014 and $2bn in 2015.

Tough Credit Compares

Credit performance at BAC is remarkable. In the last quarter, the net loss ratio on the commercial loan book of near $400bn fell below 10 basis points while that on the consumer loan book of ~$540bn was 1.0% (excluding a $144mm charge related to regulatory guidance on accounting for TDRs in the home loan portfolio) including ~3.3% on the US and non-US card portfolios. Given tight underwriting in recent years and economic improvement, the annual loss ratios will likely be lower in 2014 than 2013 before stabilizing in 2015 (see Exhibit 6).

Exhibit 6: Credit Performance at BAC

Source: Company Reports, SSR Estimates

However, the pro-cyclical behavior of reserves creates tough compares particularly in 2015 albeit moderated by the write-off of purchased credit-impaired (“PCI”) loans which is not included in reported net loss ratios. In the last quarter, for example, loan reserves declined $2bn of which $1.2bn was reserve release and $741mm was PCI write-offs; we expect these to continue at a lower rate given the PCI portfolio comprises ~$27bn of home loans. For the full year, the reserve release was ~$4.4bn and reserves nonetheless ended the year at a level covering 2.2x annual credit losses (excluding PCI write-offs). We expect this coverage level to decline over the next two years to the 1.6x of 2011 and 2012 generating a release of over $4bn in 2014 and, as loan growth picks up, flat reserves in 2015 (see Exhibit 7); this assumption creates a 26 cents swing in EPS from 2014 to 2015.

Exhibit 7: Credit Performance at BAC

Source: Company Reports, SSR Estimates

Risks in Non-Interest Income

Of course, the earnings dynamics at BAC are subject to considerable uncertainty. Volatile revenue streams, such as investment banking, gains on the equity investment portfolio, and mortgage banking, comprised 15% of BAC’s revenue of $89bn in 2013 with trading (including associated net interest income) generating a further 15%; in addition, gains on the debt securities portfolio amounted to $1.2bn. As a baseline, we assume 4-5% annual growth in the non-trading revenue streams, hold the contribution of trading to non-interest revenue flat at the 2013 level of $7bn, and reduce securities gains to $800mm and $500mm in 2014 and 2015 respectively (see Exhibit 8). In practice, we do not believe normal variation of volatile revenue-streams from these baseline estimates will meaningfully affect valuation since, absent underlying structural or environmental drivers, investors will look through.

Regulation has affected growth of the more stable non-interest revenue streams including, in particular, card income and service charges on deposits but we view the 2013 results as having reset; card income is vulnerable to a reduction in the cap on debit interchange from the present 21-24 cents/transaction either as a result of legal intervention (following Judge Leon’s ruling at the end of July that the Fed had set the cap too high) or shifting competitive dynamics (as merchants gain more control of debit routing through the MCX consortium, for example) but we expect banks, including BAC and notwithstanding the backlash in 2011 to a $5 monthly fee on debit cards, to offset this through consumer charges. The economics of signature debit, in particular, are no longer attractive to banks and we expect them to use pricing to steer consumers to PIN debit.

Exhibit 8: Non-Interest Income at BAC

Source: Company Reports, SSR Estimates

Note on Trading

Forecasting the impact of trading on GAAP results is difficult not only because of the inherent unpredictability of trading gains but also because of the accounting. A first issue is that the headline numbers for the operating performance of the trading business are the revenue reports from the Global Markets business before making the GAAP adjustment for debt valuation (reflecting the impact on the market-to-market value of certain trading liabilities based on changes in BAC credit spreads and, perversely, generating a negative item when these credit spreads tighten). In 2013, for example, there was a ~$500 million negative impact from the “DVA” adjustment (see Exhibit 10).

Exhibit 10: Detail on Trading Revenue Reported at BAC for Global Markets (“GM”) Segment

Source: Company Reports, SSR Estimates

A second issue is that trading-related revenue is spread across various line items of the consolidated income statement including, for example, net interest income, trading gains, and brokerage services (for commissions – see Exhibit 11). Hence, even accurate forecasts at the segment level for FICC and equity trading revenue are difficult to fold into a forecast income statement.

Exhibit 11: BAC – Reconciliation of Business Segment Results to Consolidated Income Statement (on FTE Basis)

Appendix A1: BAC Earnings

Appendix A2: BAC Average Balances and Credit Performance

Print Friendly, PDF & Email