Where the amplification comes from
Interest is a fixed number. If operating profit falls by a fifth, the interest bill does not follow it down — the loan agreement says what it says. Everything that operating profit loses therefore comes out of what is left for shareholders, and because that residual is smaller than EBIT to begin with, the percentage hit to shareholders is larger.
That is the whole mechanism. DFL is just the ratio of the two percentages, and it equals EBIT divided by EBIT less the fixed charges. A company with no debt has no fixed charges, so its DFL is exactly 1.00 and EPS moves point for point with EBIT. As fixed charges grow relative to EBIT, DFL rises without limit, approaching infinity as the residual approaches zero.
The amplification is symmetric, and that is the part people forget. A DFL of 2.5 turns a 10% improvement in operating profit into a 25% improvement in EPS, which is why leverage looks so attractive in a good year. It turns a 10% deterioration into a 25% deterioration for exactly the same reason. Financial leverage does not create value on its own; it redistributes an existing distribution of outcomes into a wider one.
Why preferred dividends are grossed up
Both interest and preferred dividends are fixed charges, but they sit on different sides of the tax line. Interest is deducted before tax, so a dollar of interest costs a dollar of pre-tax profit. Preferred dividends are paid out of profit that has already been taxed, so a dollar of preferred dividend requires 1 ÷ (1 − t) dollars of pre-tax profit to fund it. At a 25% tax rate, $30 of preferred dividends consumes $40 of EBIT.
Adding the two together therefore requires putting them in the same units, and the convention is pre-tax. That is why the formula's denominator carries I plus Pd ÷ (1 − t) rather than I plus Pd. Skip the gross-up and you understate the fixed charges, overstate the residual, and report a DFL that is too low — the error grows with the tax rate and with how much preferred stock is outstanding.
The denominator has a second meaning worth naming: it is the cushion. EBIT minus fixed charges is how much operating profit could disappear before common shareholders are left with nothing. The break-even EBIT — the level at which EPS hits zero — is simply the fixed charges themselves, I + Pd ÷ (1 − t). The tax rate has no effect on that break-even point for a company with no preferred stock, because taxing zero gives zero either way.
One more property makes the measure easy to use: for a fixed capital structure the relationship %ΔEPS = DFL × %ΔEBIT is exact, not just a small-change approximation, because earnings available to common is a straight line in EBIT. What changes with the size of the move is DFL itself at the new EBIT level, which is why the table under the results recomputes it at every step.
Worked example: $800M of EBIT against $200M of interest
Take the defaults, in millions: EBIT $800, interest $200, preferred dividends $30, a 25% tax rate, 120 million shares, and a 10% expected rise in EBIT.
- Gross up the preferred. 30 ÷ (1 − 0.25) = 30 ÷ 0.75 = $40M of pre-tax equivalent.
- Total fixed charges. 200 + 40 = $240M. This is also the break-even EBIT.
- DFL. 800 ÷ (800 − 240) = 800 ÷ 560 = 1.4286×.
- Earnings available to common today. (800 − 200) × 0.75 − 30 = 450 − 30 = $420M, or 420 ÷ 120 = $3.50 per share.
- Apply the 10% EBIT rise. New EBIT is $880M, so earnings to common become (880 − 200) × 0.75 − 30 = 510 − 30 = $480M, or $4.00 per share.
- Check against DFL. The EPS change is (4.00 − 3.50) ÷ 3.50 = 14.286%, which is 1.4286 × 10%. The elasticity holds exactly.
- Cushion. EBIT could fall by (800 − 240) ÷ 800 = 70% before common shareholders earn nothing.
Now run it the other way. A 10% fall in EBIT to $720M gives (720 − 200) × 0.75 − 30 = $360M, or $3.00 per share — a 14.286% decline. Same magnitude, opposite sign. The 70% cushion is the number to quote to a board, because it says how much room the business has before the capital structure starts destroying shareholder earnings rather than merely amplifying them.
How to read the number
1.00 means no financial leverage at all. Every dollar of operating profit belongs to shareholders after tax, and EPS tracks EBIT exactly.
Between 1.0 and 2.0 is a conventional range for an investment-grade industrial: fixed charges consume up to half of operating profit and a normal cyclical downturn does not threaten earnings.
Above 3.0, the cushion is under a third of EBIT. A DFL of 3.0 means fixed charges take two-thirds of operating profit, so a 33% drop in EBIT eliminates common earnings entirely. That is a defensible position for a business with genuinely stable cash flows — a regulated utility, a contracted infrastructure asset — and a dangerous one for anything cyclical.
A negative DFL means EBIT is already below break-even. The formula still returns a number, but the interpretation inverts: common shareholders are making a loss, and a rise in EBIT shrinks that loss rather than amplifying a profit. Read the earnings-to-common figure directly rather than the elasticity in that regime.
DFL is a snapshot at one EBIT level, not a property of the company. It rises as EBIT falls toward the fixed charges, which is exactly when you least want it to. Two companies with identical debt can show very different DFL simply because one is having a better year. Always look at DFL alongside interest coverage and net debt to EBITDA, which describe the debt rather than the moment.
DFL by the share of EBIT consumed by fixed charges
| Fixed charges as % of EBIT | Cushion remaining | DFL | EPS change from ±10% EBIT |
|---|---|---|---|
| 0% | 100% | 1.000× | ±10.0% |
| 10% | 90% | 1.111× | ±11.1% |
| 20% | 80% | 1.250× | ±12.5% |
| 30% | 70% | 1.429× | ±14.3% |
| 40% | 60% | 1.667× | ±16.7% |
| 50% | 50% | 2.000× | ±20.0% |
| 60% | 40% | 2.500× | ±25.0% |
| 75% | 25% | 4.000× | ±40.0% |
| 90% | 10% | 10.000× | ±100.0% |
The default case sits on the 30% row: fixed charges of $240M against $800M of EBIT, a DFL of 1.429 and a 70% cushion. Note how slowly DFL rises across the top of the table and how violently across the bottom — the last ten points of the cushion double it.
Mistakes and limits
- Forgetting to gross up preferred dividends. They are paid after tax, so they must be divided by (1 − t) before being added to interest. Omitting the gross-up understates fixed charges and DFL.
- Treating DFL as a fixed attribute of the company. It is a function of this year's EBIT. The same balance sheet produces a higher DFL in a bad year, which is when the amplification actually bites.
- Reading only the upside. The elasticity is symmetric. A structure that turns 10% of EBIT growth into 25% of EPS growth turns a 10% decline into a 25% decline.
- Ignoring capitalised operating leases. Lease payments are fixed charges too. If you are comparing a company that leases its estate with one that owns it, put the implied interest component into the interest line.
- Confusing DFL with solvency. DFL says nothing about whether the interest can actually be paid in cash. Interest coverage and debt service cover answer that; DFL only measures amplification.
- Applying it across a large EBIT change. The elasticity itself moves as EBIT moves. For a 40% downturn, recompute earnings available to common directly rather than multiplying by today's DFL.
DFL, DOL and total leverage
Financial leverage is the second of two amplifiers. The first is operating leverage: fixed operating costs — rent, salaries, depreciation — make EBIT swing more than sales do. The degree of operating leverage measures that step, from sales to EBIT. DFL measures the next step, from EBIT to EPS.
Multiply them and you get the degree of total leverage: DTL = DOL × DFL, the elasticity of EPS with respect to sales. A company with a DOL of 2.0 and a DFL of 2.0 has a DTL of 4.0, so a 10% sales shortfall becomes a 40% EPS shortfall. That product is the reason capital-intensive businesses with high operating leverage are advised to keep financial leverage modest: the two multiply rather than add, and a comfortable-looking figure on each side can combine into an uncomfortable one.
For the capital structure question underneath all of this — how much debt is appropriate at all — the relevant tools are the weighted average cost of capital, which prices the trade-off between the tax shield and financial distress, and coverage measures such as times interest earned that lenders actually write into covenants. DFL tells you what the chosen structure does to reported earnings; it does not tell you what structure to choose.
Key terms
- Fixed financial charges
- Interest plus preferred dividends grossed up for tax. Charges that do not vary with operating profit.
- Break-even EBIT
- The level of operating profit at which earnings available to common shareholders are exactly zero — numerically equal to the fixed financial charges.
- Degree of total leverage
- DOL × DFL: the percentage change in EPS for each percentage change in sales, combining operating and financial amplification.