On July 12 I briefly introduced a study by the Institute of International Finance forecasting ruin and desolation in the worst-case scenario of every single proposed banking regulation being imposed.
Additionally, my attention was drawn to a Wall Street Journal article by David Enrich, titled Studies Question Bank Capital Fears which pooh-poohed the entire notion, referring to studies by Douglas J. Elliott of the Brookings Institute and three mysteriously unnamed researchers at Harvard and UChicago.
So I’ve had a look at the paper by Douglas Elliott, titled Quantifying the Effects on Lending of Increased Capital Requirements. This paper was given a passing mention in the IIF paper, by the way.
Mr. Elliott first derives an equation giving the lower bound of the interest rate on a loan:
L*(1‐t) >= (E*re)+((D*rd)+C+A‐O)*(1‐t)
where
L = Effective interest rate on the loan, including the annualized effect of fees
t = Marginal tax rate for the bank
E = Proportion of equity backing the loan
re = Required rate of return on the marginal equity
D = Proportion of debt and deposits funding the loan, assumed to be the amount of the loan minus E
Rd = Effective marginal interest rate on D, including indirect costs of raising funds, such as from running a
branch network
C = The credit spread, equal to the probability‐weighted expected loss
A = Administrative and other expenses related to the loan
O = Other offsetting benefits to the bank of making the loan
He then assigns values to build a base case, then increases the proportion of equity while jiggling other numbers to estimate the effects on loan costs of this change.
| Base Case and Possible Future per Elliott | ||
| Variable | Base Case | Possible Future |
| Equity | 6% | 10% |
| ROE | 15% | 14% |
| Proportion Debt | 94% | 90% |
| Cost of Debt | 2.0% | 1.8% |
| Credit Spread | 1.0% | 0.95% |
| Admin | 1.5% | 1.4% |
| Offsetting Benefits | -0.5% | -0.6% |
| Marginal Tax Rate | 30% | 30% |
| Loan Rate | 5.17% | 5.37% |
He concludes that a hike in capital ratios from 6% to 10% will quite feasibly result in an increase in the loan rate of 20bp. Several variables were adjusted in this estimation:
Return on debt/deposits: Creditors ought also to be willing to drop their required returns at least modestly to reflect the lowered risk. On the deposit side, part of the adjustment might be a shift in deposit market share towards more efficient banks whose indirect cost of raising deposits is lower.
I don’t buy it. The bulk of deposits are explicitly federally insured anyway and the uninsured deposits are implicitly insured – I don’t believe uninsured depositors at US banks have yet lost a single dollar as a result of FDIC siezure. Thus, the entire 20bp reduction in the Cost of Debt will have to come from the wholesale funding markets and senior bonds. This claim will require a lot more backup than currently provided.
Credit spread: Banks ought to be able to reduce credit risk marginally by turning down less attractive loans and by imposing covenants or other features that reduce the bank’s risk of loss. The 5 basis point reduction was chosen because it appears achievable from changes in covenants and loan protections without the necessity to turn down more loans. Thus, the drop in loan supply appears unlikely to be significant.
I don’t buy it. People who “work” for banks may be a little on the otnay ootay ightbray side, if you get my drift, but I don’t believe that you can improve credit losses by 5bp simply by fiat. This claim is reminiscent of politicians who sweep into office promising more services and lower taxes, to be paid for with efficiency gains. I have no doubt but that efficiency can be improved … but show me first, OK?
Administrative costs and other benefits: These small changes would seem feasible, if banks found it necessary to alter the way they do business. This is one area where reductions in compensation could make a significant difference. Market share shifts could also account for a significant part of the change.
I don’t buy it. See above. I am particulary incensed that administrative costs are presumed to go down at the same time as extra covenants, etc, are being added to the loan terms. That doesn’t sound quite right.
So according to me, the possible future looks more like:
| Base Case and Possible Future per Elliott | ||
| Variable | Possible Future per Elliott | Possible Future per JH |
| Equity | 10% | 10% |
| ROE | 14% | 14% |
| Proportion Debt | 90% | 90% |
| Cost of Debt | 1.8% | 2.0% |
| Credit Spread | 0.95% | 1.0% |
| Admin | 1.4% | 1.5% |
| Offsetting Benefits | -0.6% | -0.5% |
| Marginal Tax Rate | 30% | 30% |
| Loan Rate | 5.37% | 5.80% |
So throwing out the more dubious changes adds 43bp on to loan cost, meaning the effect of increasing capital from 6% to 10% is not 20bp as Elliott claims, but more like 63bp according to me.
Who’s right? Who’s wrong? Who knows? There is no supporting detail in the paper that tracks the performance of the model through time. We don’t even know if the model accounts for enough of the truth to be worth-while, let alone how sticky the numbers are, or how well correllated they might be.
Anyway, according to this simple model, changing the base case in a manner with which I am more familiar, increasing capital from 6% to 10% raises the cost of a loan 63bp. 63bp! That’s a lot! Hands up everybody who doesn’t think 63bp on their mortgage is a lot!
I’m too much of an economic auto-didact to really know whether it’s strictly kosher to reverse engineer the Taylor Rule, but if we use a Taylor output gap coefficient of 0.5 then a tightening due to capital costs of 63bp means the output gap will grow by 126bp. And that’s a big number.
All in all, while the Elliott paper is interesting and provides a decent intuitive framework for a first stab at quantifying economic effects of bank capital increases, I don’t feel that there’s enough meat on these bones to justify a conclusion that we can hike bank capital and get away scot free.
And I’m still waiting for OSFI to quantify the bad effects of its insistence on high capital levels, together with the claimed good effects!