By David B. Moore and Shiv Govindan
On March 9, 2009, trading almost 65% below their December 2006 highs, both the Nasdaq Bank Index and the SNL Micro-Cap Bank & Thrift Index hit generational lows. This was not altogether surprising given the dislocation in financial markets and related economic turmoil. Aggregate nonperforming assets (NPAs) were high and rising and profitability was on the decline. As levered financial actors in a credit maelstrom, that small banks’ valuations were in the dumps was entirely understandable.
Fast forward to today and we see a much improved, albeit still challenging, fundamental picture. While many small banks continue to struggle with legacy credit issues and a shrinking net interest margin (NIM), a decent-sized subset of small banks have meaningfully reduced NPAs and are making significant strides in profitability. While most of this latter group have not yet reached “normalized” levels of profitability, they are clearly on track to earn more than their cost of capital in 2012 and for the foreseeable future. And yet the valuations of most of the banks in this group remain stuck in the mud. To wit, the Nasdaq Bank Index and SNL Micro-Cap Bank & Thrift Index were just 17.3% and 0.1%, respectively, above their March 9, 2009 lows as of the end of 3Q11. (In contrast, the S&P 500 was 70% above its March 9, 2009 low as of the end of 3Q11.) So, despite significant improvement in the fundamentals and outlook vis-à-vis the darkest days of early-2009, many high-quality small banks are trading at otherwise distressed levels. Which begs the question: When, if ever, will small-cap bank valuations return to more normalized levels?
The first issue is, of course, to define “normalized valuation levels”. In order to properly discuss this issue we must merge theory and empiricism, with the hope that the results are consistent with both.
Empiricism first: Between 1990 and 2007, the median return on tangible equity (hereafter, “ROE”) for banks with assets between $500 million and $5 billion as of year-end 2010 (almost 900 in total, thus a fairly representative sample of small banks) was 13.9%. The average trading multiple over the period for the publicly-traded subset of this group was 1.6x tangible book value. This makes sense only if the average cost of equity for the group was approximately 8.9% over the period. (In overly simplistic terms: 13.9% ROE/8.9% cost of capital = 1.6x tangible book value.) We’ll call this the “observed,” or “empirical,” cost of equity for the group.
Theory (or at least one theory) next: Over the period between 1990 and 2007 inflation averaged 2.7%, the real yield on the ten-year treasury averaged 2.9% (together equating to a nominal risk-free rate of 5.6%), and the average publicly-traded small bank had a beta of approximately 0.85. [Over-simplifying again, beta is a measure of the correlation and volatility of a security relative to that of the market.] Using the capital asset pricing model and the 5% equity risk premium observed over the 1990 – 2007 period results in a “theoretical” cost of equity of 9.9% over the period. [Cost of Equity = Risk-free rate + Beta x Equity Risk Premium; in this case, 5.6% + 0.85 x 5% = 9.9%] Thus, theory and empiricism yield similar, albeit not identical, results – in the range of 9% to 10% – for small banks’ aggregate cost of equity between 1990 and 2007.
The complication in determining the cost of equity for a bank at a given point in time is that the typical bank’s beta is fairly stable and generally below 1 during economic expansions. But during economic contractions, primarily as a result of balance sheet leverage, most bank betas increase dramatically – often more than doubling. That is, bank betas tend to be systematically understated during expansions, and dramatically overstated during contractions. For example, if a typical cycle comprises four years of expansion followed by a year of contraction, then the typical bank’s beta will spend 80% of its time below 1 and the other 20% of its time as high as 2 (or more). Thus, that bank’s beta may average slightly above 1 over a full cycle, but during the contraction its beta will skyrocket.
During the period 2008 to the present the risk-free rate has declined precipitously while bank stock volatility (and thus beta) has increased dramatically. Nevertheless – and somewhat coincidentally – the short-term cost of equity (that is, using only recent data instead of longer-term data) for most small banks is currently a bit under 9%, as the decline in the risk-free rate has more than offset the effects of higher volatility [2% + 2.25 x 3% = 8.75%; note: we have used Bernstein and Arnott’s (2002) 3% long-term equity risk premium for the recent period]. Assuming that we live in a mean-reverting financial world, the risk-free rate will rise eventually while small-bank volatility declines, such that we eventually end up in roughly the same spot where the cost of equity is concerned – the 9%-10% range – albeit with different numerical components. Clearly, however, there could be a great deal of volatility prior to such mean reversion coming to fruition.
So, now that we have a reasonable framework regarding cost of equity, we need to determine what the future holds for return on equity. In general terms, we believe that increased regulatory costs will have a modest negative impact on earnings while slightly higher capital requirements will reduce small banks’ ability to leverage. (Which, as it happens, is not a particularly controversial stance.) Consequently, that typical small bank that generated an average ROE of almost 14% between 1990 and 2007 may only be able to generate an ROE in the 11%-12% range a few years down the road (assuming the economy remains stable and NPAs continue declining). (For context, the median ROE for publicly-traded banks with assets between $500 million and $5 billion was 3.7% in 2009, 6.1% in 2010, and 9.3% in 3Q11.)
Which brings us, at last, to the issue of future valuations. If in the future the typical small bank’s cost of equity is (a normalized) 9%-10% while its return on equity hovers in the 11%-12% range, then we should expect to see average trading values in the range of 1.2x-1.3x tangible book value. While this compares favorably with current valuations (a median 0.84x tangible book at the end of 3Q11), clearly this is a significant diminishment relative to historical levels. That’s the bad news. The good news is that those banks that generate above-average ROEs should garner proportionally higher valuations. Likewise, acquisition multiples, reflecting anticipated operating efficiencies, should be proportionally higher as well. Thus, if in the “normalized” future the typical publicly-traded small bank “only” trades at 1.2x-1.3x tangible book value and is acquired for 1.4x-1.6x tangible book, it is likely that a subset of high-performing small banks will trade at over 1.6x tangible book and be acquired for over 2x tangible book.
Consequently, while we see a high likelihood of improved small bank valuations at some point in the next few years, let’s not kid ourselves – the times they are a changin’ and it is unlikely that bank valuations return to those levels we witnessed prior to the financial crisis.
[Reference: “What Risk Premium is ‘Normal’?” Arnott, Robert D. and Peter L. Bernstein. Journal of Portfolio Management. January 10, 2002.]
David B. Moore is a managing director, and Shiv Govindan president, of Resource Financial Institutions Group in New York.