LIQUIDITY

Gone are the days when one could pick up a call report and calculate a bank's liquidity using a ratio. As stated in previous editions of the BID, the simpler a calculation in banking these days, the more likely it is to be wide the mark in providing a true answer to a given problem. This is definitely the case with the liquidity ratio, which was so bad at providing an answer over the years, it expanded into 4 separate and distinct ratios. Bankers know many of these ratios by heart, which are calculated as follows: cash and unpledged marketable securities divided by total assets; non-core funding dependence; large deposits to total deposits and the loan to deposit ("LTD") ratio. The problem with many of these calculations has been driven by the very evolution of banking. Thirty years ago, 91% of banks used core deposits to fund the majority of their assets. Today, that percentage has dropped to about 55%. Household savings have been sliding over that same time period and money has become more and more fungible. Given a shortage in core funding and strong asset growth, banks have shifted to other sources of funding. Consider, for instance, that over the past 10Y, FHLB borrowings alone have climbed a whopping 500%. Before we go too far into the past with our discussion, let's fully understand the weaknesses of each liquidity calculation. We begin by examining cash and unpledged marketable securities divided by total assets. There are many problems with this ratio, but the biggest is that it ignores the loan portfolio entirely. As with securities, loans can also be pledged as collateral for borrowing, a fact missed by this ratio. This and other problems with the ratio reduce the value of it as a proxy for bank liquidity. Next, we look at non-core funding dependence. Here, bankers take the difference between non-core liabilities and short-term investments and divide them by long-term assets. The thought behind this one was that short-term investments were better funded by non-core sources than long-term assets. Again, the problems with this ratio are many. To begin, consider that the calculation uses stated maturity date, not expected repayment date. Given the optionality embedded in the loan portfolio these days, using final maturity date is practically worthless. In addition, longer-term funding is considered to be non-core, which flies directly in the face of the concept of matched funding. The third calculation, large deposits to total deposits, was intended to be a proxy for funding stability. The problem with this ratio is that the size of a deposit is not necessarily a risk driver. In fact, some large deposits are quite stable, while others can be volatile. The intricacies of deposit relationships, seasonality and other factors must also be considered. Finally, we review the LTD ratio. This ratio assumes that not only that loans are illiquid, but also that any funding shortages would be made up for by the sale of securities. The primary problem with this ratio is that loan portfolios are often used for liquidity (i.e. participations) and it ignores funding through repo, the FHLB, brokered CDs, or Fed Funds. As one can see, using ratios to measure and manage liquidity is a bad idea. Instead, bankers should understand projected cashflow, borrowing capacity, earnings projections, asset growth, interest rate expectations, funding costs, economic conditions, contingent events, stresses, marketability and use of collateral, lines of authority and responsibility and a realistic (tested) view of the length of time needed (timing and general ability) to liquefy various assets. As funding sources evolve, so too must bankers in order to effectively measure, monitor and control liquidity risk. Given the fundamental shift away from traditional sources of liquidity, the time has come for independent banks to reassess how liquidity risk is calculated