"The lesson of history is that you do not get a sustained economic recovery as long as the financial system is in crisis."
It’s once again earnings season and a great deal of attention will be focused on financials. Over the past three months, the equity market values of most of the largest universal banks have traded off as investors have started to appreciate that the party is ending in terms of new mortgage originations driven by refinance transactions. As I noted in the last post, the guidance from all of the big banks is decidedly negative for Q3 because of the prospective decline in revenue and transaction volumes in mortgages.
While refinance transactions are falling rapidly, mortgage loan purchases volumes are not growing nearly enough to make up for the drop in overall volumes. The chart below shows the total loan originations, refinance and purchase volumes for all lenders from the Mortgage Bankers Association through Q1 2013:
Close your eyes and imagine what this chart will look like next year. Looking at the banking industry as a whole, the mortgage story is likely to dominate the headlines next month during earnings season -- even if the financial media wants to ignore the implication for the “housing recovery.” A lot of analysts want to believe that the relatively modest rise in interest rates since the bottom last summer is the culprit in terms of falling mortgage loan volumes, but my view is that three factors – declining affordability, a stagnant job market and flat to down consumer income – are the structural factors behind the anemic demand for mortgage loans, particularly mortgages for home purchases.
While banks are happy to make loans, especially jumbos, to existing customers for refinance transactions, the new Basel III rules and GSE lending standards make it problematic for banks to move down the risk curve, especially if doing so takes us outside the agency bucket patrolled by the New Calvinists at the Consumer Finance Protection Bureau. Loans that are not “qualified mortgages” that can qualify for a federal guarantee are very costly for banks, both in terms of capital costs and charges for liquidity, MSRs, etc. The chart below shows bank mortgage portfolios for first and second liens, and sales and securitization volume, for all FDIC insured banks.
As you can see from the chart, the total retained portfolio of real estate loans held by US banks has dropped about 20% since 2007, from ~ $5 trillion to $4 trillion today. The major area of shrinkage has been in 1-4 family loans, while second liens have also been shrinking slowly. Sales and securitization of second liens is very small and was not included in the chart. The series for 1-4 family loan sales and securitizations is also falling, again owing to the secular decline in agency volumes and other factors. As the watering hole shrinks, the GSEs and TBTF banks will consume one another in vicious competition. By the way, has anybody noticed that Freddie Mac is discounting its loan guarantee fees in competition with Fannie Mae? We’ll come back to that soon…
Jumbo loans were the only part of the mortgage complex to show growth in 1H 2013, up 17% YOY according to Inside Mortgage Finance. Jumbo originations were about $115 billion in 1H 2013.
The sharp inflection point in 2009, of note, is attributable to the new FASB rules regarding off-balance sheet accounting for special purposes entities. While banks still sell most of their origination volumes into the agency market, the portion being retained on balance sheet is slowly shrinking.
So given that the Case-Shiller national survey of home prices is up 12% YOY, how does one interpret the decline in bank financing for home purchases over the same period? The easy answer is that about half of the gains reflected in Case-Shiller over the past couple years are attributable to the gradual resolution of bank REO and the disappearance of the spread between distressed home sales and voluntary home sales. About half of all home purchases were cash during 2012. But when you consider that the ex-REO gain in Case-Shiller is about half of that 12% figure YOY and that bank credit underlying the housing market has been shrinking all the while, how does that make you feel about the future prospects for home price appreciation (HPA)? Hold that thought.
In terms of industry revenue and earnings, the general is more important than the particular. For example, the increase in Q2 2013 earnings was largely driven by increases in non-interest income and reserve releases. Trading income also spiked. There is not a lot of organic revenue growth in the US banking industry today. Thus as some of the larger players have been guiding down on mortgage lending volumes, they have also warned of potential losses on the mortgage line because there is nothing available to take up the slack. The FDIC’s excellent Quarterly Banking Profile summarizes the situation:
“Noninterest income was $6.7 billion (11.1 percent) higher than in second quarter 2012. Income from trading rose by $5.1 billion (238.3 percent) compared with a year ago, when the industry reported a net loss on credit derivatives. Net gains on sales of loans and other assets were $1.9 billion (63.7 percent) above the level of a year earlier. For the third quarter in a row and fourth time in the last five quarters, net interest income posted a year-over-year decline, falling by $1.8 billion (1.7 percent) as interest income from loans and other investments declined faster than interest expense on deposits and other liabilities. Banks set aside $8.6 billion in provisions for loan losses during the quarter, a $5.6 billion (39.6 percent) reduction from a year earlier. This is the lowest quarterly loss provision for the industry since third quarter 2006, when quarterly provisions totaled $7.6 billion. Total noninterest expense was $1.4 billion (1.4 percent) lower than in second quarter 2012, when industry expenses were elevated by restructuring charges.”
So the clear message to take away from the Q2 2013 data on the US banking industry is 1) cost cutting in terms of operations, 2) lower loan loss provisions and 3) increased non-interest fees are the key factors in terms of revenue drivers. There is no real visibility in terms of revenue growth. In Q3, however, the sharp drop in mortgage volumes is going to upset the carefully scripted ballet that has kept large bank earnings within an acceptable range for the Sell Side analyst and media communities.
Given that mortgage origination and sale has been the dominant revenue line item for many of the largest banks over the past ten years, you would think that the financial media would be all over this story. After all, JPM has actually guided to an operating loss in mortgage in Q3-Q4. But no, instead we talk about pointless government litigation against bank shareholders and the London Whale. Go figure.
Keep in mind that Q2 earnings were also helped by a 10% increase in the “fair value” of mortgage servicing right or MSRs, a non-cash adjustment that goes right to income thanks to the idiocy of fair value accounting. Just as the market for non-performing loans was a little fluffy in Q1 of this year, the market for MSRs is also showing a bit of foam right now. The real question is whether these markups will need to be adjusted, again, as and when the excitement subsides. Normally public companies don’t toy with the valuation of intangibles except at year-end.
So when we actually start the Q3 earnings cycle for financials, watch for the word “surprise” in a lot of news reports and analyst opinions. Nobody seems to want to take notice of the very public guidance coming from some of the largest names in the banking complex because of what it implies for housing. But just to show you that God has a sense of humor, Bank of America and Citi have actually outperformed their asset peers in the TBTF group over the last three months. Hey, that’s what we need, an index comprised of TBTF banks. Be a useful surrogate for the credit quality of the United States.
See you at Americatyst 2013 in Austin TX next week.