Price on Value
July, 1999
All Earnings Are Not Created Equal
John Price, Ph.D.
Perhaps "all men are created equal", but the same cannot be said of
earnings. Investors hover around earnings releases and predictions like moths around
a candle. And for good reason. For short-term investors it is vital to know whether
earnings for the next quarter or two are going to have any surprises in either direction.
For longer term investors a central goal is to predict the growth rate of earnings over a
decade or more.
It may be that predictions of future earnings can be made more precise by going beyond
simple quantity to look at their quality and components.
There is plenty of evidence that the growth rate of earnings is mean-reverting. This is
a statistical term implying that if earnings have been growing at a high rate compared to
the industry in general, their growth is likely to move towards the industry norm.
Similarly, growth rates that are lower than the norm tend to rise.
Research by Richard Sloan reported in The Accounting Review (1996) shows that
there are indicators within the earnings figures that help to determine the speed of this
mean reversion. In a nutshell, it is the amount of cash in the earnings figures. If it is
high, then it is more likely that the earnings figures remain at their current level.
Conversely, if it is low, the mean reversion occurs more quickly. You can take advantage
of this by giving more significance to the cash component of earnings.
Before giving more details, it is useful to look at the precursor of this idea in the
work of Benjamin Graham in which he placed varying degrees of importance on different
items in the balance sheet. In Security Analysis, Graham wrote that the first rule
in calculating liquidating value is "that liabilities are real but the assets are of
questionable value." The dependability of the assets were listed by Graham as:
Asset Class |
% of Face Value |
| Cash assets |
100% |
Receivables |
75-90% |
Inventories |
50-75% |
Fixed assets |
1-50% |
In other words, all assets are not created equal when it
comes to determining an operational level of book value. They have, according to Graham,
different levels of quality.
There is now good evidence that earnings also have different levels of quality. First
of all, are earnings repeatable? An increase in earnings usually comes from one of two
sourcesgreater sales or lower costs. With a few exceptions, sales can continue to
grow year after year at a substantial rate. In contrast, cost cutting cannot be continued
year-after-year at the same rate.
Consider Gillette. In the 1988 annual report we read that net sales have increased by
7% over the past five years while net income has increased by 13%. Also it has a return on
sales of 10.7%. But if sales and income are to continue increasing by these amounts each
year, in five years return on sales would be 14% and in ten years it would be 18.5%.
Although these returns are not impossible, they are highly unlikely since the average
return on sales in the toiletries/cosmetics sector is only 8%.
In other words, although it is reasonable that sales will continue to grow at their
past rate, it is doubtful that the same will be true for earnings.
Another gremlin with earnings figures is the sale of assets. When this happens,
earnings can lurch upwards. Clearly such earnings are even less repeatable.
The proportion of earnings in cash as compared to book-keeping entries is another
touchstone for the quality of earnings. A simple estimate of this is to follow
Grahams advice and caste your eye over the list of current assets. If there is a
drop in cash and cash equivalents compared to an increase in all the other entries such as
receivables and inventories, then you would be advised to tread carefully.
In a speech last year to the New York University Center for Law and Business, Arthur
Levitt, the chairman of the SEC, said that earnings were like a bottle of fine wine: you
wouldnt pop its cork before it was ready. "But some companies are doing this
with their revenue," he continued. "Recognizing it before a sale was
complete."
A more sophisticated approach is given by Sloan in which he measures the ratio of
earnings in cash and earnings as accruals. Roughly, accruals are that part of the net
revenues that are recognized in the accounts but which are not in cash. Under accounting
based on accruals, revenues are recognized when goods or services are provided which
typically precedes cash collection. Sloan showed that the higher the proportion of the
cash component of earnings to the accrual component then the greater is the persistence of
earnings performance.
In his conclusion, Sloan writes, "Perhaps the results in this paper are simply
evidence of a normal return to an active investment strategy based on financial statement
analysis." So, despite this careful analysis involving thousands of companies over 30
years, it may be that in the end it comes down to what every value investor already
knowslook at the fundamentals. To this, I am certain Benjamin Graham would add
"while always allowing for a margin of safety."
Next months article "The Joy of Gain Versus the Pain
of Loss" will discuss some of the ways that we make decisions. Are we as rational
as we think we are? |