GTI Corporation: Using Z-Score to Manage a Financial Turnaround
Corporate Financial Distress, Restructuring, and Bankruptcy by Edward I. Altman, Edith Hotchkiss, and Wei Wang (Wiley, ISBN 978-1-119-48180-5)
Previous: Distress Prediction for External Analysts | Next: Sovereign Default Risk and Z-Metrics
People kept asking Altman the same question: “Your model says we’re going to fail. Now what?” For years, his answer was basically “hire a turnaround specialist.” That satisfied nobody. Chapter 12 tells the story that changed his mind.
In 1975, a CEO named Jim LaFleur took the Z-Score model and ran it in reverse. Instead of using it to predict bankruptcy, he used it to prevent one. The company was GTI Corporation, a parts and equipment manufacturer listed on the American Stock Exchange. The case is over 40 years old. Altman says it is just as relevant today.
Passive vs. Active Model Use
Most predictive models assume the observer does not influence the outcome. Analysts score a company, assign a probability, and move on. LaFleur treated the Z-Score as a management dashboard. Before every major decision, his team simulated the impact on the five ratios. If a move would not improve the score, they reconsidered.
That is a fundamentally different relationship with a model. The prediction becomes a target to beat, not a forecast to accept.
How Bad Was GTI in 1975?
LaFleur joined the board after reading about the Z-Score in Boardroom Reports. When he took over as CEO in mid-1975, the numbers were brutal:
- Working capital down $6 million in six months
- Retained earnings down $2 million
- $2 million net loss
- Net worth fell from $6.2 million to $4.4 million
- Market value of equity cut in half
- Sales down 50%
The preliminary Z-Score came in at 0.7. With accurate figures, it dropped to 0.38. That put GTI near the median score for bankrupt firms (about zero) with a bond-rating equivalent around CCC- or worse.
Retrospective analysis showed the score had been falling for at least two years, even during periods when reported profits looked fine. The model saw what the income statement hid.
The Core Strategy: Kill Underutilized Assets
LaFleur identified the root problem quickly. GTI’s total assets had ballooned while the numerators in the Z-Score ratios (working capital, retained earnings, EBIT, equity value, sales) did not keep up. Bloated assets dragged every ratio down.
The fix: find assets not earning their keep, sell them, and use the proceeds to pay down debt. Selling assets cuts the denominator (total assets) in all five ratios simultaneously. Using proceeds to reduce debt improves the equity-to-liabilities ratio (X4). Both moves push the Z-Score up.
Inventory was the first target. Returned goods sat uncounted. Work-in-process far exceeded what sales justified. Excess inventory was liquidated, sometimes at scrap value. Corporate staff went from 32 people to 6. Two unprofitable West Coast plants were cut to skeleton crews within 10 days. Capital spending froze.
One Sale Changed Everything
Cost cuts and inventory liquidation helped, but the turnaround accelerated in late 1976 when GTI sold its crystal base product line. The product matrix analysis showed the line was marginally profitable but not complementary to GTI’s other products. It needed heavy capital investment to stay competitive. Demand was fading.
The sale generated cash to pay down debt. Total assets and debt both fell. The Z-Score jumped from under 1.0 to 2.95 in a single transaction, nearly reaching the safe zone above 2.99.
LaFleur had felt the company was recovering before the outside world noticed. This sale confirmed it.
The Numbers Over Time
| Period | EPS | Z-Score Direction |
|---|---|---|
| 1975 H1 | -$1.27 | 0.38 (near bankruptcy) |
| 1976 | $0.28 | Rising |
| 1977-78 | $0.15 to -$0.29 | Still climbing |
| 1979 | $0.70 | Safe zone |
| 1981 | - | 8.8 |
Debt-to-equity fell from 128% to 30% over five years. By 1981, the debt-to-market-equity ratio was under 10%. GTI was acquired by a Scandinavian firm in 1992, but by then it had been a financially sound, conservatively managed company for over a decade.
What LaFleur Did Right
A few things stand out:
- He simulated before acting. Every major decision ran through the Z-Score first.
- He attacked the balance sheet, not just the P&L. The previous management chased sales growth with debt. LaFleur shrank the asset base.
- He sold emotionally difficult assets anyway. The ceramic capacitor division stayed open longer than it should have because the technology was interesting. When he finally closed it, the Z-Score rose despite an earnings hit.
- He involved employees. The people running the machines knew where waste lived.
My Take
This is my favorite chapter in the book so far. Not because the math is novel. Because it shows a model doing something models almost never do: changing the outcome they predict.
Altman did not design the Z-Score as a management tool. LaFleur repurposed it. That is a lesson for anyone who builds analytics: your users will find applications you never imagined, if the model is simple enough to manipulate.
The GTI story also exposes what most turnaround advice misses. Consultants talk about vision and culture. LaFleur talked about five ratios and whether a decision moved them in the right direction. Unsexy. Effective.
The closing note about Chapter 22 (repeat bankruptcy) matters too. Altman says the Z-Score should be a barometer after restructuring, not just before. If you emerge from Chapter 11 still scoring like a distressed company, you are probably heading back. GTI proved the opposite path works.
Previous: Distress Prediction for External Analysts | Next: Sovereign Default Risk and Z-Metrics