Palm Beach | We think 2017 will be remembered as the Year of the Bubbles. Everywhere you look, whether its stocks or real estate or even overt acts of fraud like bitcoin, the value of fiat paper dollars measured in prices for other “assets” is falling. Crypto currencies, to be clear, are more a class of felony than investable assets, but the crypto games provide supply for demand in an age of scarcity engineered by the central banks.
During a visit to the Atlanta Fed last week, we had a fascinating dinner with a group of institutional investors. Like many metros around the US, Atlanta real estate is booming after years of post-crisis lethargy and ample amounts of monetary gasoline from the FOMC. Our friend Dick Hardy organized the dinner. He noted, going back to crypto, that bitcoin is really about a shrinking float of available tokens, an ingenious aspect of the bitcoin scheme. But in the world of bank credit, scarcity and abundance exist simultaneously.
Bank earnings in 2018 are likely to continue to rise with asset returns, and expenses are likely to fall as regulatory changes ripple through the world of banks and non-banks alike. The big wild cards are an inverted yield curve in the US and an economic slowdown in China, as we discussed in our interview with Lee Miller of China Beige Book. But in terms of valuations for financials, current equity and asset returns for US banks remain significantly below pre-crisis levels. Chart 1 below shows asset and equity returns for all US banks.
For the past five years, there has been a bull market in residential and commercial real estate, albeit for properties that are decidedly up-market. With the increased cost of regulation, the minimum threshold for a residential mortgage that somebody actually wants to service has risen proportionately. If the mortgage has a unpaid principal balance of say $300k or less, it is just marginally profitable for many servicers – especially those located in CA.
We note in this regard that Walter Investment Corp (WAC) just filed for bankruptcy court protection. Other non-bank players in the residential mortgage space are struggling. During our discussion last week in Atlanta, MBA chief economist Mike Fratantoni reported that non-bank lenders managed to get profit margins back up to about 40bp in Q3 ‘17 from single digits in the first half of the year, but depositories fared far worse in the MBA survey. A decade ago, residential lending returns were over 2%.
Consider high-touch First Republic Bank (FRC), the San Francisco based lender that focuses on managing assets for high-income clientele and making jumbo mortgages for same. As of Q3 ’17, the gross spread on all of the bank’s real estate loans – which is 80% of FRC’s loan book – was just over 3%. The overall 3.1% yield for FRC’s entire loan book is half a standard deviation below its peers, according to the TBS bank Monitor. Chart 2 below shows FRC’s gross loan spread vs its asset peers.
Source: FDIC/TBS Bank Monitor
FRC has never particularly focused on smaller mortgages, preferring the well-bid world of jumbo loans, which the bank sells into securitizations managed by the likes of Redwood Trust (RWT) with servicing retained. Most other banks large and small have fled the low end of the residential mortgage market and focus primarily on tri-coastal jumbos over $1 million. Chase, Wells Fargo (WFC) and Bank America (BAC) are super competitive on jumbos over $1 million in urban markets. The larger institutions win business by offering APRs and other terms that are well-below that of conforming loans half that size – and barely make money. They typically keep prime jumbo loans in portfolio.
Having escaped the below-prime world of FHA mortgages, overall bank loan credit in 1-4 family mortgages is pristine. The net charge-off rate in Q3 ’17 was just 0.04%, largely because home prices are rising so fast thanks to the Yellen Inflation that banks are having a hard time losing money when that rare mortgage default event occurs. In Chart 3 below, note that past due 1-4 family loans remain stubbornly high at 2.6% in Q3 ’17, this due to the backlog of foreclosures that remain in the judicial states of the Northeast. The glacial pace of foreclosures in judicial states, which often exceeds 1,000 days from default to resolution, is just one aspect of the cost of “consumer protection” for MBS investors.
The positive impact of rising home prices on bank credit is shown in Chart 4. In Q3 ’17, loss-given default (LGD) in 1-4s reach a new low of just 24%, the lowest observation for this metric since at least 1990. The 30-year average LGD for 1-4s is 66%, a fact that will perhaps be of note to our friends at the Board of Governors in Washington. The plummeting LGD for residential mortgages owned by banks illustrate very graphically how the actions of the Fed have boosted home prices and greatly advantaged home sellers.
Since banks avoid the bottom third of the US mortgage market in terms of credit quality, the credit outlook for banks is quite positive – but we still expect to see defaults slowly rise from the current low levels. Seeing a 24% LGD is a skew, an outlier. Because of the sharp supply shortage of 1-4 family homes as well as affordable apartments in many markets, we do not expect to see prices decline appreciably as the Fed ends QE. But as we told the audience in Atlanta on Friday, we don’t expect to see mortgage interest rates rising because of the dearth of duration in the bond market.
Last week in The Institutional Risk Analyst, we talked about how the fact of the Fed’s ownership of $4 trillion in Treasury paper and MBS has taken away upward pressure on bond yields. Since none of the global central banks that collectively own $20 trillion plus in debt and equity hedge their positions, there is no selling pressure to push bond prices lower and yields higher.
Banks and other fixed income investors are trapped in the world of “lower for longer” so long as the central banks retain their bloated securities holdings. Indeed, as we predicted during the discussion at the FRB Atlanta, we expect to see an inverted yield curve in Q1 ’18.
Last week, Peter Cecchini, chief market strategist at Cantor Fitzgerald, called the flattening yield curve “the most important thing to have a clear idea about now.” This is especially true for banks and other financials, which have surged past the S&P 500 and other equity market benchmarks in the collective madness surrounding stocks in the runup to the tax cutting legislation. Sure, net-interest margins have been rising for banks, however we believe that the prospect of a flat or inverted yield curve will give investors and FOMC members reason for pause.
Chart 5 illustrates the recent upturn in bank interest income even as interest expenses have risen far more slowly. So far, banks seem to have managed to keep hungry depositors at bay as yields have risen from 2015 lows, but the Fed is still effectively transferring $80 billion per quarter from depositors to banks. Note how wide the net interest margin grew in 2009 when the Fed slashed rates but yields on earning assets were still relatively high compared with today.
As we’ve noted in previous missives, bank returns on the $16 trillion or so in earning assets are still quite subdued at just shy of 80bps. And the market for new bank loans in sectors such as C&I and commercial real estate remain extremely competitive for larger banks, putting an effective cap on loan yields. So unless bond spreads expand and loan yields actually rise from current levels – something we think is unlikely – the bullish improvement in bank interest earnings may slow.
More, if as we suspect the yield curve inverts next year, the FOMC may need to rethink its schedule for benchmark rate increases. Bank credit metrics look quite good at present – too good really. Negative net loss rates for multifamily loans and construction & development exposures remind us that the FOMC has greatly skewed the world of credit – in some cases by several ratings notches. This anomaly will eventually be reversed, revealing tens of billions worth of mispriced exposures on the books of US banks.
As Chart 6 below suggests, the cost of credit for construction and development loans in the US remains badly skewed and has been negative since 2015. The degree of downward deviation from the 30-year average LGD of 60% suggests that the adjustment could be far more severe than the 2007 financial crisis – if and when a more general deflation of asset prices occurs. But it remains to be seen whether asset prices can adjust in the near term.
So the good news is that bank earnings likely will to continue to improve with relatively low credit costs, but a flat Treasury yield curve may change that trend. Asset and equity returns for US banks remain 1/3 below pre-2008 levels. Loan growth will probably continue to decelerate from the torrid levels of 2015 and 2016. Whether or not anyone on the FOMC gets the joke in the near term and starts to sell MBS and long-dated Treasury bonds is perhaps the most important question facing bank investors as 2017 comes to a close.