With record-low interest rates and a volatile stock market, many midtier financial institutions are quietly taking on more risk to maintain performance.
Unless proper analysis accompanies this risk-reward trade-off, they may unwittingly be setting themselves up for an earnings shortfall.
The primary risk, from both the asset and liability sides of the balance sheet, is that when interest rates rise, the institutions will be whipsawed by an interest rate mismatch that may harm results.
The problem is especially acute for midtier financial institutions, because their asset and liability modeling are typically less rigorous and disciplined than those of larger institutions. Banking regulators routinely report that many institutions either don't do enough asset-liability modeling or fail to conduct the necessary evaluations to determine the effectiveness of their models.
Compounding the problem is that midtier institutions lack access to derivative products that most large institutions use for basic hedging strategies.
The pressure to perform is driving bankers toward more credit and interest rate risks. In the best case, if rates and chargeoffs remain steady, the strategy will work well and produce a healthy return on equity. However, the risk is considerable - unless an institution's financial modeling can provide the intelligence to make informed decisions.
ASSET AND LIABILITY RISKS
In today's market, understanding the true risks on both the asset and liability side of the balance sheet are key, particularly with the interest rate climate likely to change appreciably over the next 12 to 18 months.
The most common problem stems from bankers fighting to replace the refinancings that have streamed in. As repayments reduce loan holdings, banks have scrambled to maintain net interest margins.
To stave off competition, banks have either consciously or unconsciously taken on more credit or interest rate risk than they normally would. In particular, savvy commercial borrowers are attempting to lock in long-term financing that takes advantage of the 40-year low in rates. Instead of going out two to three years - as many were doing just 18 months ago - borrowers are now seeking to tie up low-cost loans for seven years or longer.
To offset the risk, banks making these loans have either tapped Federal Home Loan Bank advances or have saturated the certificate of deposit market. In both cases, banks have decreased the amount of asset-liability flexibility they will have in the future.
But satisfying restless borrowers is only half the challenge. In trying to keep $100,000 depositors happy, many bankers realize that they must pay above 1%, the unofficial Maginot line of banking. In some competitive markets, bankers have had to offer up to 3% to retain a relationship.
The cost of these deposits is putting more of a crimp in margins and forcing bankers to take on additional credit or interest rate risk.
Also, liability risks are worsening as bankers are lulled into a false sense of complacency. With few attractive alternatives, many smaller depositors have been content to keep their money in bank accounts, either waiting for rates to rise or using the accounts as an alternative to the equity markets. Banks have fallen into the trap of lumping these deposits in with their core deposit balances, and have underestimated the sensitivity of that money.
Banks that have lent on a longer-term basis to increase margins will often reallocate core deposits into a longer duration for modeling purposes in an effort to stay in compliance.
History has shown that when the Federal Reserve moves interest rates, it often does so in a decisive manner to stay ahead of inflation. This being the case, increases of 50 to 100 basis points in a single quarter are not uncommon. Any rapid movement in rates will quickly expose banks' balance sheets to asset and liability mismatches, and those mismatches will directly impact earnings.
Should the Federal Reserve march steadily back to a 6% target for overnight federal funds, many of a bank's assets could be below its costs of funds. To make matters worse, prepayments would slow, causing the duration of assets to increase, while liabilities would shorten as depositors go back into the equity markets or seek the yields of longer-term investments.
In scenarios such as these, a bank's asset-liability model pays for itself in a matter of minutes.
ACCURATE MODELING
Given where interest rates are today, bankers would be well-advised to validate the assumptions in their existing financial models and then perform the necessary back-testing to determine how well those assumptions fare - and how well they are likely to do when interest rates rise.
In validating the model, bankers should first determine whether the model has the power to predict the right direction. For example, if a model forecasts a 12% earnings increase, but they actually fall by 5%, the model needs to be adjusted immediately.
Assuming that the model accurately predicts direction, it should then be evaluated on a quantitative basis for how close it comes to matching reality. As a general rule of thumb, banks should strive to produce models with a variance of less than 30% of actual results.
Bankers can get a glimpse of how their models are likely to perform when rates rise by comparing their current model to the period from July 1999 to June 2000. During that time the Fed raised rates a total of 175 basis points in a coordinated series of hikes. When the Fed starts tightening again, there are likely to be notable similarities to that period. Sound assumptions supported by back-testing are critical to validating the accuracy of a model over time.
Everyone remembers the interest rate debacles that produced the savings and loan crisis in the 1980s and led some money market mutual funds to break the $1 share price in the early 1990s. Midtier financial institutions can go a long way toward sidestepping these mistakes by reviewing their models and making midcourse adjustments.
If nothing else, having confidence in an asset-liability model will lead to more accurate management decisions and the ability to better optimize that risk and reward equation.
