Federal reserve chair Janet Yellen’s defense of the benefits of regulation last week in Jackson Hole probably killed her chances for reappointment, but the more pressing reason to see Yellen return to the private sector is visible in the U.S. real estate market.
Yellen and her colleagues have created large bubbles in many asset classes from residential homes to commercial real estate to construction lending. As in the 2000s, this latest bout of asset price inflation will not end well for banks or investors.
What do the most recent bank loan portfolio data from the Federal Deposit Insurance Corp., for the second quarter of 2017, say about asset prices and inflation? For some quarters now, the credit statistics for the $16 trillion asset banking system have been too good to be true, in some cases suggesting that credit events have no cost. The last time this circumstance existed was the mid-2000s, when several large mortgage banks were reporting a negative cost — that is, a profit — from default events.
The same real estate market dynamic that allows growing numbers of Americans to take cash out of their homes is depressing the cost of loan defaults to half-century lows. Even faced with this rather striking situation, our faithful public servants on the Federal Open Market Committee (FOMC) can actually stand up in public and say that inflation is too low. The skews in the credit world are so large that some banks are actually earning a profit on recoveries after a loan balance is repaid in full.
First let’s examine credit trends for 1-4 family mortgages, a $2.4 trillion asset class for U.S. banks. Loss given default (LGD), a fancy way of expressing net charge-offs, shows the average loss for such mortgages. At the quarter’s end, the LGD for this asset class was just 24%, the lowest loss rate net of recoveries since at least 1990. Last quarter, the volume of defaults on these mortgages fell below $1 billion, or less than a tenth of a percent of total loans.
Residential assets prices are quite high, as reflected by the high recovery value — 76% — implied by the 24% LGD in the second quarter. Since 1990 the average loss rate after a default for 1-4 family loans is 67%, so it seems reasonable to ask when U.S. home prices will adjust downward. How you feel about that depends upon whether you view the extraordinary home price inflation seen since 2012 as being permanent and thus immune to mean reversion.
The situation in the world of construction lending is even more profound, with LGDs well into negative territory for the first time since the 1990s. In the second quarter, LGD on those few construction loans that actually defaulted was negative 94%. Given that C&D loans tend to be mostly multifamily paper and have loan-to-value ratios around 50%, when you see a bank reporting such unusual profits on defaulted loans it suggests that the value of the real estate has basically doubled since the bank made the loan. Note that the downward move in LGD coincides with the end of quantitative easing by the FOMC.
Of note, home equity lines of credit (HELOCs) are showing behavior similar to that of 1st lien mortgages, which typically stand in front of HELOCs in the credit stack. LGD for HELOCs reached a mere 31% in the second quarter, implying that banks are recovering almost 70% of the value of a loan when a borrower default occurs. The 25-year average LGD for HELOCs is 65%, illustrating that Yellen and her colleagues on the FOMC have literally turned the world of real estate credit on its head. The value of the collateral backing HELOCs has surged since 2012.
Next let’s move to the $780 billion in credit card loans held by U.S. banks, an asset class that has seen recent growth after years of flat-to-down portfolio levels. The interesting thing about credit card loans is that they are totally unsecured. Thanks to the generosity of chair Yellen and the other members of the FOMC, credit card loss rates have fallen dramatically since 2008.
Note that default and non-current rates are both turning up after the years of irrational easing by the FOMC. Credit card charge-off rates are above those levels for non-current loans, the opposite of that relationship for most other loan types held by U.S. banks. This is because, being unsecured, these credits tend to be charged off before they have an opportunity to be classified as non-current. LGDs for credit card loans were at 83% for the second quarter, the lowest since the mid-2000s.
Here’s a key question: how much future default risk has been buried under the comforting blanket of low interest rates? The average rate of net charge-offs for credit cards is almost three quarters of a point above current levels, again begging the question as to when we shall revert to the mean. The answer to that question will have a significant impact on bank earnings and banks’ ability to return excess capital to shareholders.
Finally, let’s take a look at the $1.9 trillion portfolio of commercial and industrial (C&I) loans, traditionally one of the most important indicators of future U.S. economic growth. Defaults and non-current rates are both below the averages going back to the 1980s, while credit spreads are as compressed as ever over that same time period. Note that net charge-offs for all bank-owned C&I loans are below those for non-current loans, which may end up being worked out or restructured short of a formal default.
While net-charge offs for C&I loans are relatively normal compared to the real estate-related asset types, loss rates measured by LGD have been climbing since 2015. More important, new loan production rates (as well as sales) are falling such that the portfolio of bank owned C&I loans is no longer growing very much. This suggests that the U.S. economy is slowing and demand for credit is therefore on the wane.
It’s important to note that the charge-offs and recoveries reported in each period by FDIC insured banks are disparate events. A recovery reported today might be related to a loan charged off three years ago. But the key element of price is reflected in the aggregate data reported by each bank, so that a rising recovery rate/falling LGD strongly suggests that prices in the underlying market are quite frothy. Just as in the 2000s and the 1990s, the Fed again has stoked an asset price bubble in real estate that may lead to significant losses for banks and bond investors should prices correct.
Whoever gets the top job at the Fed, the FOMC must live with the balance sheet and market conditions served up by Chair Yellen and her predecessor, Ben Bernanke. While the Fed’s initial focus on narrowing credit spreads was correct, the FOMC should have stopped after QE1 ended in 2010. Instead, concerned that banks were not lending, the Fed continued to buy securities.
The Fed has kept the pedal to the metal, repeating the errors of the 2000s by fueling a credit driven bubble in residential real estate. This time around, the bubble is more focused on affluent areas of the country, but the result is likely to be more tears.
BTW, we’re rooting for Kevin Warsh as the dark horse candidate for the Fed job in the event that White House chief of staff Gary Cohn decides to stay put. But the biggest challenge facing Yellen’s successor as Fed Chair is having the courage to admit that inflating asset bubbles does not create jobs or prosperity, just future financial crises.
Christopher Whalen is chairman of Whalen Global Advisors and a former senior managing director for Kroll Bond Rating Agency. This article was initially published in Whalen’s online publication, The Institutional Risk Analyst. It is reprinted here with Whalen’s permission.