Leverage Considerations in Financial Analysis

Leverage Considerations in Financial Analysis

Leverage relates to the variable and fixed cost mix inherent in firms’ cost structures. Convention typically recognizes two types of leverage – operating and financial. “Operating” leverage relates to fixed operating costs (e.g., rent) while “financial” leverage relates to fixed financing costs (e.g., interest). Functionally these two types of leverage influence net profit in the same way. This note introduces the concepts of operating and financial leverage and how they relate to changes in common profitability measures. Techniques for fixed and variable cost estimation using publicly available accounting data are also briefly reviewed.

Operating and Financial Leverage

The leverage effects on a firm are “relative” in that the impact on net profitability is dependent on the current profit level. Appreciating the importance of leverageimplies understanding the sensitivity of earnings to estimated changes in sales. All else equal, for a highly levered firm a given sales change will have a higher impact on profits. A simple example illustrates this point:[1]

ABC / XYZ
Fixed costs / $ 0 / $ 200
Variable costs/Sales / 80% / 40%
Assets / $1,000 / $1,000

Case 1: Assume no debt financing (e.g., liabilities=$0, equity=$1,000)

ABC / XYZ
Scenario / A / B / C / A / B / C
Sales / $ 400 / $ 600 / $ 800 / $ 400 / $ 600 / $ 800
Variable costs / 320 / 480 / 640 / 160 / 240 / 320
Contribution margin / 80 / 120 / 160 / 240 / 360 / 480
Fixed costs / 0 / 0 / 0 / 200 / 200 / 200
Operating income / $ 80 / $ 120 / $ 160 / $ 40 / $ 160 / $ 280
Interest expense / 0 / 0 / 0 / 0 / 0 / 0
Net income / $ 80 / $ 120 / $ 160 / $ 40 / $ 160 / $ 280
Profit margin / 20% / 20% / 20% / 10 % / 27 % / 35 %
ROA / 8 % / 12 % / 16 % / 4 % / 16 % / 28 %
ROE / 8 % / 12 % / 16 % / 4 % / 16 % / 28 %
OLE / 1.00 / 1.00 / 1.00 / 6.00 / 2.25 / 1.71

Observe for Company ABC that for any percentage change in sales there is the same percentage change in total profits and the profitability ratios. Company ABC has zero operating leverage. Company XYZ, however, is levered as it has fixed costs in its cost structure. For a given increase in sales, net income will, therefore, increase at a faster rate. To quantify that rate some analysts usea ratio labeled the “operating leverage effect” (OLE):

OLE = Contribution Margin  Operating Income.

The intuition behind this ratio is that for a firm the difference between the contribution margin and operating income is its fixed costs, and thus the ratio reflects the relative size of fixed costs to operating income. When OLE is greater than one, operating leverage exists.As can be seen from the illustration OLE will vary based on the company’scurrent sales and profitability state. For instance, from scenarios B to C operating leveragedecreases from 2.25 to 1.71, reflecting a lower impact of fixed costs for higher levels of sales and contributions margin. OLE can be used to interpret forecasted sales assumptions as follows:

If XYZ finds itself with sales of $600, a 33% decrease in sales will result in a 75% (= 2.2533%) decrease in earnings, ROA and ROE.

What about financial leverage (i.e., debt in the capital structure)? The concepts of operating leverage and financial leverage are effectively the same to equity holders as the interest component can be considered just anotherfixed cost borne by the firm. Case 2 assumes the same data as Case 1, with assets instead one-half financed with debt.

Case 2: Assume 50% debt financing (e.g., liabilities and equity =$500), interest rate = 5%

ABC / XYZ
Scenario / A / B / C / A / B / C
Sales / $ 400 / $ 600 / $ 800 / $ 400 / $ 600 / $ 800
Variable costs / 320 / 480 / 640 / 160 / 240 / 320
Contribution margin / 80 / 120 / 160 / 240 / 360 / 480
Fixed costs / 0 / 0 / 0 / 200 / 200 / 200
Operating income / $ 80 / $ 120 / $ 160 / $ 40 / $ 160 / $ 280
Interest expense / 25 / 25 / 25 / 25 / 25 / 25
Net income / $ 55 / $ 95 / $ 135 / $ 15 / $ 135 / $ 255
ROA / 8 % / 12 % / 16 % / 4 % / 16 % / 28 %
ROE / 11 % / 19 % / 27 % / 3 % / 27 % / 51%
OLE / 1.00 / 1.00 / 1.00 / 6.00 / 2.25 / 1.71
FLE / 1.45 / 1.26 / 1.19 / 2.67 / 1.19 / 1.10
TLE / 1.45 / 1.26 / 1.19 / 16.00 / 2.67 / 1.88

The effects of financial leverage have no impact on ROA from what was reported under Case 1, because this return metric is measured before the effects of interest and capital structure. For equity holders, however, interest is a fixed cost like any other that must be covered before a return can be earned. The “financial leverage effect” (FLE) is a ratio that reflects this burden and is

FLE = Operating Income Net Income

As with operating leverage, FLE will vary conditional on the currentprofitability state. For instance, in scenarios B and C for Company ABC the financial leverage effect is 1.26 and1.19, respectively. FLE is interpretable as follows:

If ABCfinds itself with sales of $600, a 33% increase in sales will result in a 42% (1.2633%) increase in earnings and ROE.

The total leverage effect (TLE) captures the net effect of both OLE and FLE, and is simply the product of these two components. Analysts wishing to quantify leverage effects must be able to categorize all costs as either fixed or variable, a difficult and frequently impossible task.

Fixed and Variable Cost Estimation

Consideration of the level of fixed costs in a firm’s cost structure is important for all firms because sales are rarely static – a firm with high leverage will be more sensitive to growth estimates in sales. The techniques for leverage estimation, while simplistic, result from data limitations given companies do not segregate fixed costs from variable costs. Regardless of the estimation method chosen, one should first and foremost never ignore known changes in the operating environment that would impact the level of fixed costs (e.g., plant expansion or contraction, work force reductions). As with all prospective analysis the MD&A is a good starting point to identify these macro changes, wherein management will often willingly describecostbehavior and future operating plans.

Estimation of fixed costs is, unfortunately, a very noisy process. In the absence of direct knowledge of fixed costs, analysts often rely on observations of historical levels of common size expense ratios and account change patterns, and assess whether those relations will recur. Two methods are generally employed, (i) account by account analysis, or (ii) linear estimation with quarterly or annual data. The latter takes the form of either two-period estimation or regression estimation. Regardless of the approach taken, an important assumption of any fixed cost estimation technique is that the cost structure does not change over the period used in the estimation. Note that this is often an unreasonable assumption for regression analysis as this approach requires a long time-series of data. The two-period and account-by-account estimations have,consequently, been found to perform equally well as regression.

1. Account by account – This approach employs common size income statements coupled with knowledge of which costs are generally fixed and which are variable. If a cost varies with sales consistently each year, it is considered to be variable. If a cost remains unchanged as sales changes, it is fixed. Examples of fixed costs would be depreciation expense provided capital expenditures are not required to fund an increase in sales, rent and lease expense.

1. Linear estimation – This approach is also crude, but in practice works well when the firm has reached a level of stability. The method requires the use of historical data to estimate a linear cost function, based on either the most recent two periods of data or with regression analysis using multiple periods of data.

Variable definitions:

S = Sales; FC = Fixed costs;

VC =Variable costs; VC% = VC/Sales;

TC = Total costs; NI = Net income;

CM = Contribution margin; CM% = CM/Sales;

n = Unit volume; p = Sales price;

v = Variable costs per unit

Two-Period Estimation

Two-period estimation is a technique that simply plots the two most recent years of data to identify a variable costs percentage. This percentage is then used to split a total cost line item into its variable and fixed cost components. It is based on the following fundamental relations:

TC = FC + VC = FC + (VC% × S)

S – VC – FC = NI

S - VC - FC = NI

S - VC = NI; (because by definition FC = 0)

This relation implies that TC/S = VC/S (again, FC = 0)

Thus, TC/S uses known data (total expenses and total revenue) and yields the variable cost percentage (VC%).

With VC%, simply compute FC directly.

FC = TC – (VC% x S)

Regression Estimation

This technique is similar to two-period estimation, but considers a longer time series of costs. To estimate the fixed and variable cost mix using regression one can take a variety of approaches, using available data for net income (NI), total costs [TC, sales (S), and possibly unit volume (n)].

Note the multiple ways to represent net income and total costs:

(1)NI = S – VC – TC = (p×n) – (v×n) – FC = -FC + (p-v)n

(2)TC = VC + FC = FC + (v×n)

(3)TC = FC + (VC% × S)

Recall that simple regressionstake the general form:

with  the estimated intercept, and β the estimated slope of a linear fit of the data.

Estimation of equations (1), (2) and (3) yield an intercept,, interpretable as the level of fixed costs. The estimated slope coefficient, β, is either the contribution margin per unit (under equation 1), the variable cost per unit (equation 2), or the variable cost percentage (equation 3).

Which regression equation should one use if all data is available? Many consider equation (2) to be superior, as variable costs vary directly with the level of output. Using (1) or (3) incorporates price (p), and the introduction of this price will introduce an additional source of measurement error. Equation (3), however, is the most common technique given it does not rely on the frequently unavailable volume measure. Every income statement has total revenue (“S” that maps into X) and total expenses (“TC” that maps into Y). The regression of TC on S yields output of estimated fixed cost dollars () and the variable cost percentage (β).

1

[1] Adapted from White, et.al. (1999), The Analysis and Use of Financial Statements.