Every year our Gov't needs to pose better questions when it comes to our money. It is ridiculous to be asked, "Do you want to donate $3 to the Presidential Election Campaign Fund?" The Federal Gov't doesn't think to ask if you want to earmark $3 to help children, the impoverished or the sick, just politicians. Major politicians don't need any more money and it's a travesty that candidates like Lyndon LaRouche receive a penny. LaRouche has indicated he believes the Queen of England is an international drug trafficker and that Microsoft is a terrorist organization and we don't want our money going to him.
For the past year, we asked the wrong question during our research, when we starting to look at the performance of 2k banks over the past 10Ys. We asked, "What is the best loan mix to have to produce superior earnings?" It turns out, like Presidential candidates, there is no perfect mix. All loans can produce a bank ROE of 15% or greater as long as times are good. That is a significant finding, as it argues for diversification and generating loans that the community demands. However, our problem was that we asked the wrong question.
The question should have been, "What loan types produce a superior ROE with the lowest amount of risk?" When we equalized loan performance against contribution to earnings and earnings volatility, a completely different picture emerged. While some loans can produce a superior ROE, they can also produce a negative ROE.
One of the worst performing risk-adjusted loan categories over the last 10Ys has been multifamily, followed by residential real estate. In other words, banks with concentrations in these two areas, most likely have a higher chance to see swings in earnings. As we drilled down into the data, the main reason this seems to emerge is that these categories are often mispriced when it comes to risk. This is another reason why banks need to understand where risk is coming from, how it impacts the loan value and areas of future performance concern. Our Loan Pricing Model helps connect actual loan risk by zip code to return. Multifamily for instance, has historically produced larger and more frequent delinquencies than say office lending, but most banks price multifamily at smaller spreads to Libor. This is the exact opposite of what should occur. In addition, many banks fail to take into account interest rate risk when generating fixed rate loans in both multifamily and residential, which argues for banks to take advantage of our BLP program whereby we take on the interest rate risk. Finally, the 3rd reason that these loans produce greater than average earnings volatility is due to higher and more volatile overhead and origination costs. Both residential and multifamily have a fair amount of administrative and production costs that eat into earnings.
Our analysis finds that the best loan type to have on your books is agriculture backed by real estate. This loan category outperformed all others over the last 10Ys, making it the loan category that best contributes to stable earnings. Again, this is not to say that defaults are lower, because they are actually higher than both multifamily and residential mortgage lending. This is to say that banks that generate agricultural loans backed by real estate tend to price these loans at spreads that generate excess returns to take into account credit, capital and expenses.
If you want to see how other loan types compare, be sure to attend our Executive Management Conference in May in S.F. https://cbc.webcorr.com/PCBBPublicApps/CFO2009RegistrationForm.jsp
In the meantime, let's hope the IRS comes around and asks better questions with regard to our tax dollars.