If one wished to isolate a single statistic that summed up the elemental health of any bank, it would not be an accounting magnitude like the return on equity or the return on assets.
Rather, it would be a market-driven measure that foreshadows changes in underlying solvency. Such a measure would set forth the probability within a given period (say, one year or five years) that the market value of the bank's assets will decline to the level where that value just equals the book total of its liabilities, which is, of course, the point of default.
This probability of insolvency or expected default frequency is a figure readily available to banks and nonbanks that are publicly traded. It was pioneered and is now routinely calculated by KMV Corp. of San Francisco.
A recent review of the statistic gives cause for grave concern. The expected default frequency of the median megabank (with assets of $100 billion or more) for the ensuing year has been rising, almost uninterruptedly, since February and began to spurt upward alarmingly beginning in July. This institution enjoyed a default probability of only 7 chances in 10,000 in February; by October, that probability had tripled, to 22 chances in 10,000.
The situation at the nation's second-tier banks (assets of $30 billion to $100 billion) is much less grave, though still worrisome. The median institution in this category saw its default likelihood rise from only 5 basis points in February to 12 basis points in October. That represents a higher default likelihood than at any time since November 1994.
Default probabilities like these may seem modest to the unsophisticated, but senior bankers know differently. To qualify for an AA rating, which is necessary if the institution wants to do an appreciable amount of derivatives business, a bank cannot afford to run a default likelihood consistently above 5 basis points. That's because the default frequency historically associated with AA-rated companies ranges from 2 to 9 basis points, depending on the stage of the credit cycle.
By contrast, the range of default frequencies associated with BBB-rated firms is considerably wider. Analysis reveals, however, that at the current stage of the cycle, the frequency should approximate 17 to 19 basis points. It follows that unless the median $100 billion-plus bank can quickly lower its default likelihood from the level of 22 basis points recorded in October, it may no longer qualify for an investment grade.
We do not believe that the nation's biggest banks will or should lose their investment grades. Indeed, we are basically bullish on the banking industry. Still, many first- as well as second-tier institutions are entering an exceedingly difficult period-one that will require them to reexamine and upgrade their risk management skills.
Why has the median megabank been moving closer to default?
An increase in the probability of default can be spurred by one or both of two factors: a rise in the volatility of asset values (of the downward type) and an increase in financial leverage.
Self-evidently, as asset values trend downward, the likelihood that any given amount of equity support will be sufficient to protect the institution diminishes, thereby shortening its distance to default. An increase in financial leverage has the same impact since it reduces the capacity of the bank's equity base to sustain any given amount of asset- value depreciation.
The rise in the default likelihood of the median megabank since February has been for the most part the result of increasing leverage rather than enhanced asset volatility.
In the KMV model, leverage is expressed in market terms and is represented by the ratio of the nonequity claims to the market value of assets. At the typical giant bank, every dollar of assets supported nonequity claims of 85 cents in February; by October the total of claims had topped 90 cents.
For most of this period, the volatility of asset values remained constant at 4%. (That is, one standard deviation of asset values amounted to plus or minus 4%.) In midsummer, however, asset volatility began moving up from 4% to 5%, which exacerbated the negative impact of rising leverage.
At the second-tier banks, the rise in leverage was somewhat more muted than at the giant institutions. This was in part the result of a leveling off in the growth of book liabilities after May.
Whereas the total of book liabilities increased sharply at the biggest banks, it remained virtually unchanged at the second-tier institutions-a circumstance suggesting that most of the loan growth in this mature stage of the economic cycle is being preempted by the largest banking entities.
It should be noted that since, in the KMV model, leverage is a market- determined phenomenon, it is powerfully affected by what happens in the stock market. The market value of bank assets-the denominator of the leverage formula-cannot be computed directly. (Nor could it be even if market prices for all bank loans were readily available.) But the market value of the assets is definitionally equal to the market value of the liabilities, and this latter magnitude can be computed.
KMV starts with changes in the market value of the equity, uses these changes to adjust the market value of the nontraded bond debt (employing an option-pricing approach), and finally adds in the book total of the current liabilities to arrive at an up-to-date value for the bank's liabilities, which is also the value of its assets.
Thus, a drop in equity values is associated with a fall in the value of the noncurrent debt, culminating, other things being equal, in a decline in asset values and thus an increase in the institution's leverage.
If the difficulties of the megabanks and the second-tier institutions are traceable primarily to rising leverage and, perhaps incipiently, to enhanced asset volatility, then the remedies would seem to be straightforward. To reduce leverage: Raise equity and hold fewer assets. To curb downward asset volatility: Get better-quality assets and diversify those you already have.
Quite obviously, these deceptively simple solutions are far from simple to implement. At times, lack of implementation stems from differences of opinion.
For example, many bankers would still balk at raising equity in a soft market. But the status of a bank's market-to-book ratio is, or should be, irrelevant to the decision of whether or not to raise equity.
True, the marketplace has beaten up bank stock values, but the same process that savaged share prices greatly widened the spread between risky and riskless assets. So while bank spreads may not have been high enough to cover customer risks in mid-1998, this may no longer be true in early 1999. Banks can and should raise equity if they can persuade the marketplace that rising spreads make it likely that the risk-adjusted returns on new equity funds will exceed the opportunity cost of these funds.
However, most of the difficulties involved in implementing solutions to rising default probabilities stem not from differences of opinion over what to do, but from organizational and cultural impediments to doing what most agree has to be done. This is especially true with respect to measures needed to dampen asset volatility. These revolve around three basic initiatives:
Improving the quality and the consistency of the grading, capitalization, and pricing of both funded and unfunded loan extensions.
Eliminating obstacles to increased loan sales and purchases or, alternatively, to the sale and purchase not of the actual loans but of the risks embedded in them.
Changing the composition of the bank's bond portfolio to offset risk concentrations in the loan portfolio.
In subsequent articles we will be elaborating on these three essential initiatives, which, by materially lowering default probabilities, can banish the threat of punishing credit downgrades. We will also be analyzing more of the trends signalled by the KMV model. u
