Goldman Sachs And Citi Looking Even Cheaper On Strong DFAST Results

| About: Goldman Sachs (GS)
This article is now exclusive for PRO subscribers.


The benchmarking power of DFAST is sadly ignored as the tests play second fiddle to CCAR.

Goldman Sachs and Citi did quite well on DFAST, and the test is tangible evidence that these banks' price to books should expand.

Too many big banks are still trading below book-value post-crisis. Is it really justified when certain of these banks are both doing well on DFAST and have 10% ROEs?

Does the equity market already price in DFAST results? Not always.

I picked over the Dodd-Frank Act Stress Tests (DFAST) results released last Thursday looking for meaningful relationships between the results, which are solid proxies for risk, and the subject banks' valuations. Traditionally, we think of 'higher risk' as 'higher reward' and thus, arguably higher valuations. Not so with big banks anymore. Lower risk, on balance, is tending to get more rewarded for banks -- but only if the ROEs are there. For many banks, the ROEs are not there.

There is some degree of relation between lower risk sensitivity (as demonstrated by each bank's DFAST results) and these banks' valuations, but ROEs remain considerably more dominant in driving banks' (still) weak price-to-book values, many of which seem to remain crippled by the overhang of the financial crisis.

It's a bit unfortunate that the information in DFAST's 'equal-footing on capital distributions' is probably quite under-utilized by banks and the market. Every bank has again passed and there is no information contained in the DFAST about return of capital plans or the qualitative assessments of risk management -- all of which is included in CCAR. Yet, the DFAST tests are a wealth of standardized data that could be more key to shaking off the post-crisis overhang and can be helpful informing valuation discounts.

Let's turn to the relationship between ROEs (5y medians) and the Price/Book values (as of Friday June 24, post-Brexit) of the 33 DFAST banks. The graph below shows that ROEs have a 72% coefficient of determination (r-squared) with Price/Book ratios.

Data Source: YCharts

The first observation is that the fit is pretty good. The red points are the eight G-SIBs. Achieving higher ROEs likely has A LOT to do with achieving higher Price/Book values -- not surprisingly. And if we look at the 10% ROE point, we see that DFAST banks should be valued at about a 1.0x price-to-book -- that too seems fair.

Let's cut out of the data the top-five highest ROE banks (processors and card companies), and five of the weaker ROE foreign banks. The remaining banks are more of the 'core' U.S. banks. Below we have:

Data Source: YCharts

The power of historical ROEs to explain Price/Book values drops considerably, to an r-squared of just 35%. Without American Express (NYSE:AXP), Bank of NY Mellon (NYSE:BK), Discover Financial (NYSE:DFS), State Street (NYSE:STT), Deutsche Bank Trust (NYSE:DB), MUFG America Holdings, BBVA Compass, BMO Financial (NYSE:BMO), TD Group US Holdings (NYSE:TD) and HSBC North America Holdings (NYSE:HSBC), the valuation differences are less explained by ROEs. I show this smaller, but more 'core' bank group to get a cleaner comp group. Can we use the overlay of DFAST results to see who might be more mis-priced? We can.

Banks below the line -- especially the ones well below the line, could be taking on risks that help explain why their Price/Book ratios are below expectations. Goldman Sachs (NYSE: GS) and Citi (NYSE: C) stand out. Should these banks trade well below the ROE regression line in light of DFAST results?

Goldman had a CET1 minimum under the severely adverse scenario of 8.4% -- a healthy 390bps above the 4.5% pass/fail threshold. Its average change in its four capital ratios was -35.1%, not nearly as severe the much worse -54.3% drop in 2015's severely adverse test. GS's projected nine-quarter pre-provision net revenue (PPNR) mysteriously was a whopping five-fold above the level of the 2015 test. Regardless why that was, things are generally going right for GS in the 2016 DFAST.

Similar observations can be made for Citi, which has done very well in DFAST over the past two years. It's minimum CET1 was 9.2% in the severely adverse scenario this year (470 bps above 4.5% pass/fail threshold), and it's losses as a percent of beginning RWAs have improved in each of the past two years.

Thus, if we believe that the DFAST is a high quality stress test that accurately reflects the risk sensitivity of large U.S. banks, then should GS and C trade so far below book, or so far below the regression line? No, I don't believe so. I believe the stress tests are arguments for seeing some improvements in the Price/Book values of these two banks. Examples like GS and C make me believe that some DFAST results are NOT priced into the market.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.