The ratio worked on one bank, the distance to 100 per cent with losses charged as problem loans grow, and the bond mark the published figure leaves out.
The Texas ratio is a bank's nonperforming assets, meaning nonaccrual loans, loans 90 days or more past due and foreclosed real estate, divided by its tangible common equity plus its loan loss reserve. On an illustrative bank with 126,000 of nonperforming assets against 600,000 of tangible common equity and a 65,000 allowance, it is 18.95 per cent. The warning level is 100 per cent; here nonperforming assets would have to rise 4.06 times, to 9.8 per cent of loans, to reach it once the losses on them are charged.
Worked in full in Bank Management by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The ratio was developed by analysts at RBC Capital Markets studying the Texas bank failures of the 1980s, and later applied to New England: banks that crossed 100 per cent failed far more often than those below it. It survives because it asks one blunt question. If every problem asset became a loss, would the bank's own loss-absorbing money cover it?
| Item | Amount |
|---|---|
| Total loans | 5,200,000 |
| Nonaccrual loans | 96,000 |
| Loans 90+ days past due, still accruing | 18,000 |
| Other real estate owned (foreclosed) | 12,000 |
| Total shareholders' equity | 690,000 |
| Goodwill and other intangibles | 90,000 |
| Allowance for loan and lease losses | 65,000 |
| Unrealised loss on held-to-maturity securities, after tax | 250,000 |
Step 1: the numerator. Add the three kinds of problem asset: 96,000 + 18,000 + 12,000 = 126,000, or 2.42 per cent of loans. Exclude the portion of any loan guaranteed by a government agency, which is not at risk.
Step 2: tangible common equity. Total equity less goodwill, other intangibles and any preferred stock: 690,000 − 90,000 = 600,000. Intangibles go because they cannot absorb a loss; preferred goes because it ranks above the common holders who are meant to absorb it. The full calculation is in how to calculate a bank's TCE ratio.
Step 3: add the reserve. The allowance is money already set aside for losses, so it sits in the denominator: 600,000 + 65,000 = 665,000.
Texas ratio = (nonaccrual + 90 days past due + OREO) ÷ (tangible common equity + loan loss reserve)
= 126,000 ÷ 665,000 = 18.95%
In Excel: =(Nonaccrual+PD90+OREO)/(Equity-Intangibles-Preferred+ALLL)
The simple answer is to divide 665,000 by 126,000: nonperforming assets could rise 5.28 times. That overstates the distance, because new problem loans do not just sit in the numerator. They lose money, and the loss comes out of the reserve and then out of equity, shrinking the denominator at the same time.
If each new nonperforming dollar loses 40 per cent, the level N at which the ratio reaches 100 per cent solves N = 665,000 − 0.40 × (N − 126,000):
N = (665,000 + 0.40 × 126,000) ÷ 1.40 = 511,000
4.06 times today's level, 9.8 per cent of loans
| Multiple | NPAs | Loss charged | Denominator | Texas ratio |
|---|---|---|---|---|
| x1 | 126,000 | 0 | 665,000 | 18.9% |
| x2 | 252,000 | 50,400 | 614,600 | 41.0% |
| x3 | 378,000 | 100,800 | 564,200 | 67.0% |
| x4 | 504,000 | 151,200 | 513,800 | 98.1% |
| Loss severity | NPAs at 100% | Multiple of today | Share of loans |
|---|---|---|---|
| 0% | 665,000 | x5.28 | 12.8% |
| 25% | 557,200 | x4.42 | 10.7% |
| 40% | 511,000 | x4.06 | 9.8% |
| 60% | 462,875 | x3.67 | 8.9% |
The definition matters too. On total equity rather than tangible, the same bank reads 16.7 per cent. Leaving out the loans 90 days past due, as some screens do, gives 16.2 per cent. Neither is wrong as arithmetic, but a ratio quoted without its definition cannot be compared with anyone else's.
Tangible common equity, as published, carries held-to-maturity securities at amortised cost. This bank has 250,000 of unrealised loss on them after tax. Take it and tangible common equity is 350,000, the Texas ratio 30.4 per cent, and the nonperforming assets that would take it to 100 per cent fall to 332,429: 2.64 times today, 6.4 per cent of loans. In the stress table, tripling the problem assets gives 67.0 per cent on the published figures and 120.3 per cent with the bond mark taken. A bank whose credit book looks sound can sit much closer to the line than the published ratio suggests, which is the theme of the book's four readings of one balance sheet.
The common mistake is to read the Texas ratio as a credit measure only, and a static one. It is a race between problem assets and loss-absorbing capital, and both sides move: problem loans grow and consume the reserve as they do. Running the ratio forward with a loss severity on new problem assets, and with the securities marks the regulatory figures leave out, turns a reassuring 18.95 per cent into a distance to failure that can be argued about in numbers.
The other quiet error is in the numerator. Loans that have been restructured for a borrower in difficulty but are still accruing are not in any of the three categories, yet they default at far higher rates than the rest of the book. A bank that modifies problem loans early can keep its Texas ratio low for several quarters. Ask for modified loans separately and run the ratio with and without them.
Add nonaccrual loans, loans 90 days past due and foreclosed property, divide by tangible common equity plus the reserve, then solve for the problem-asset level that reaches 100 per cent with losses charged as they arise. Here: 18.95 per cent today, 100 per cent at 511,000 of problem assets, or 332,429 once the held-to-maturity mark is recognised. The four readings of capital, bridged to the dollar, are in the free workbook for this case.
Lower is better, and 100 per cent is the traditional warning line, the point at which problem assets equal the bank's loss-absorbing capital. The illustrative bank's 18.95 per cent is comfortable on its face, but with 40 per cent losses on new problem loans it would reach 100 per cent at 511,000 of nonperforming assets, 9.8 per cent of loans.
Because the reserve is capital already set aside to absorb loan losses, so it stands alongside tangible equity as a cushion. In the example the 65,000 allowance lifts the denominator from 600,000 to 665,000. As problem loans grow and are charged off, the reserve is consumed first, which is why the ratio rises faster than the numerator alone.
Not as usually published: tangible common equity carries held-to-maturity bonds at amortised cost. Deducting this bank's 250,000 after-tax unrealised loss cuts tangible equity to 350,000 and lifts the ratio from 18.95 to 30.4 per cent. Tripling problem assets then gives 120.3 per cent rather than 67.0.
This article is one calculation from Bank Management. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: regulation, compliance and banking → · All 1340 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.