Yet viewed through a different lens, the cause of the loan growth may not be that troubling.
In August, I identified three possible reasons for loan growth sharply outpacing the overall rate of growth in the economy: banks are increasing their risk profile at the wrong time; banks are looking to boost short-term earnings even if that means long-term returns decline; or, finally, borrowers are enticed by cheap credit, which is not necessarily a problem unless commerce becomes overwhelmed by leverage.
But now I would like to pinpoint a fourth possible reason. When using long-term metrics to track the loan growth, one finds lending overall could simply be reverting to the pre-crisis mean. In other words, with the credit markets ceasing activity in 2008-10, lending is potentially just coming back to more historical levels. Yet, if growth is due to this seemingly benign reason, it still has enormous implications to bankers and regulators the next time there is a serious downturn in the economy.
Let me explain further.
The Federal Deposit Insurance Corp. reported on Aug. 30 that loans grew 6.7% year over year as of the second quarter. During the same period GDP grew roughly 1%. Historically, bank profits eventually suffer when loan growth gets too far ahead of the overall health of the economy.
For more background, it is necessary to revisit the "Great Panic" years of 2008 to 2010. This is how former Bank of England Gov. Mervyn King referred to the financial crisis in his 2016 book "The End of Alchemy." He suggests an already fragile financial world was made much worse by widespread panic.
FDIC bank data seems to support his view. A close examination of the quarterly FDIC data shows that bank lending froze up beginning in the second half of 2008. Total bank loans peaked on June 30, 2008, at $8 trillion. Lending did not return to that level until the second quarter of 2014.
While banks effectively stopped lending during the Great Panic, loan-loss provisions hit unprecedented highs. From 1984 to 2006, the ratio of provisions to assets — on an annualized basis — averaged 0.53%. During the Great Panic, the ratio reached a high point of 1.95% and averaged 1.52% from 2008-10. In 2014, the ratio fell below 0.20% for the first time in history.
One big reason loan-loss provisions fell so significantly beginning in 2011 is because so many loans were charged off from 2007 to 2009. In fact, the historic record indicates too many loans were charged off. What's the proof of that?
During the 12 quarters ending December 2009, aggregate loan-loss provisions for the industry exceeded chargeoffs by $184 billion. That number is whopping considering that the best four consecutive quarters of earnings for U.S. banks is $163.9 billion. Having overreacted to the Great Panic — perhaps because of political influence — bankers and regulators reversed course from 2010 to 2015 when chargeoffs exceeded provisions by $121 billion.
Of course, provisions weren't bankers' only focus during the Great Panic. Capital-happy regulators and academics forced banks to bolster capital ratios. Rather than seek dilutive outside capital, the vast majority of banks slowed lending and improved their ratios by shrinking risky assets, i.e., loans.
The net effect of these actions hurt the U.S. economy as banks failed to bring necessary capital to growing businesses.
All this is background for the fourth possible explanation as to why U.S. bank loan growth seems so far out of whack with GDP.
Because the years leading to, during, and after the Great Panic were so extraordinary, it may be more appropriate in 2016 to evaluate loan growth on a five-year trend. Here's where the data gets interesting.
Going back to 1984, the five-year growth rate of U.S. bank loans has averaged 22%. What is the five-year loan growth for U.S. banks at the end of the second quarter 2016? Answer: 22%.
In a classic case of reversion to the mean, the five-year growth rate averaged 46% from 2005 through 2007, compared to -2% from 2011 to 2013. Add the two time periods together and the average is 23%.
The current disconnect between loan growth and GDP may be explained by the severe slowdown in loans during and immediately subsequent to the Great Panic. In essence, bank lending in the U.S. is now getting back on its long-term natural growth trajectory.
Just to be clear, if I were a bank director or chief risk officer I would still be worried by today's loan growth numbers. Aggregating U.S.-wide loan numbers is very useful for evaluating one form of systemic risk, but a more detailed sifting of loan numbers is required to ensure banks maintain a disciplined understanding of risk and return.
The devil is always in the detail. Thankfully, the regulatory community is fully aware that multifamily lending and construction and development lending are growing, respectively, 3.4 and 4.6 times faster than their 30-year averages. For sure, those growth rates vary tremendously across the country, suggesting banks in certain parts of the country need to be more alert than others.