US Banks and Margins

Banks have been troubled investments since 2007. The housing crisis and resulting aftermath destroyed the equity value for most banks. A few didn’t survive or were absorbed by others.

After the crisis, banks had new difficulties: increased capital requirements, limitations on business activities, accounting changes, and increased oversight.

Additionally, fintechs with better technology and customer acquisition capabilities have entered the space.

Stock performance has been weak. Many US banks trade at single digit earnings multiples, or even below tangible book value. Some banks could even be considered net-nets, depending on how generous you are on asset liquidity and debt levels.

Banks also currently have limits on their dividend payouts and are restricted from repurchasing shares. The Fed has reserved the right to evaluate this policy on a quarterly basis.

Finally, and perhaps most importantly, net interest margins (NIMs), at US banks have been on a 10+ year slide. Net Interest Margin is the spread of what the bank earns on its assets (i.e. loans) against its costs (i.e. deposits). This metric is the key driver which determines their profitability.

One theory for this reduction in NIM the past decade has been the decline in interest rates themselves. One of the bear cases on US banks is that NIMs will continue to decline as interest rates go lower, similar to what happened in Japan.

But is there a relationship between interest rates and NIM? In the below chart, we compare net interest margins for US banks compared to 10 year government rates since 1984.

Source: FRED

Interest rates have been declining throughout the entire 35 year period. The ten year government rates dropped by over 1000 basis points. Meanwhile the decline in US NIMs was about 70 basis points.

Not exactly a 1-1 move.

Source: FRED

In Japan, NIMs are indeed much lower. In the late 1990s, they averaged around 1.2% and are currently around 0.8%. We only have data through 2017, but NIMs dropped about 40 basis points since the mid-1990s. At the same time, 10 year interest rates dropped by 300 basis to go negative. Again, not a 1-1 move.

NIMs in Japan have always been lower than the US. This is similar to how returns on equity in Japan have just always been lower. Japan is a fundamentally different place in terms of business culture. For example, Japanese banks return far less capital via share repurchases.

From this analysis, it seems that NIMs have moved down about 10-15% of the move in the rates themselves. And while the absolute levels may vary between countries, there is no obvious reason why NIMs in the US should drop to Japanese levels.

All of this data occurred in periods of falling interest rates. At this point, you are probably wondering what happens to NIMs when interest rates rise?

Fortunately Wells Fargo has annual reports from all way back to 1969 posted on their website. Here is their NIM compared to the US 10 year:

Source: Wells Fargo annual reports, FRED

Just as before, there is little correlation between NIM and interest rates. Wells Fargo increased its net interest margin through the mid 1990s, before embarking the slow slide to today’s lower levels. From the early 1980s through the 1990s, its NIMs were increasing as rates were falling.

Overall, the correlation between government rates and bank NIMs is weak. If anything, the direction of NIM appears to be more correlated to economic activity.

So will NIMs in the US continue to go down? Possibly. There is increased competition from fintechs and regulatory oversight is not lessening. If they do continue their decline, the decline will take a long time.

Future stock returns of US banks will depend on your view of competition and the US economy longer-term. Given the way banks are currently priced, it seems like a decent bet to me.