In my last article, I explained why investing in stocks with low Price-to-Earnings (P/E) ratios tended to deliver good performance. To summarize, if companies earn similar amounts in the future as in the past, those with low P/E ratios are likely to be trading below their intrinsic values. But, I also mentioned several problems with using the ratio, one of which centers around debt.

Recall the following definition of the P/E ratio:

price_to_earnings = price_per_share * number_of_shares_outstanding / net_income

A high P/E company can “magically” lower its ratio by borrowing money to repurchase its shares. Doing so would lower both `number_of_shares_outstanding` and `net_income,` but if the company could borrow at reasonable rates, the first number would decrease much more than the second. Assuming the `price_per_share` is unaffected, some simple math shows that the P/E would drop. Nothing about the business has changed; the company didn’t gain any new customers or build a new factory. But with this transaction – “Poof!” – the company suddenly became more attractive as an investment.

P/E’s deficiency stems from only considering shareholders as the only source of financing. When a company builds a new factory, for instance, it’ll generally use a mix of shareholders’ monies and debt, and profits generated from the factory will pay both shareholders and lenders. The difference between these types of financiers is that 1) lenders get priority, and 2) they only receive a fixed sum. Whatever is left goes to the shareholders. So, what if we treated lenders like a special type of shareholder with the privileges we mentioned? That’s the philosophy behind the EV/EBIT ratio, which is defined as follows:

shareholder_financing = price_per_share * number_of_shares_outstanding
lender_financing = sum(all_debt)
enterprise_value = shareholder_financing + lender_financing - cash
ebit = net_profit + interest_expense + tax_expense
ev_ebit = enterprise_value / ebit

Enterprise Value, which ‘EV’ stands for, is the combined financing provided by shareholders and lenders. Notice that it subtracts cash from the amount. That’s because the company can theoretically use cash to reduce the amount financed by shareholders and lenders via share or debt repurchases.

EBIT stands for ‘earnings before interest and taxes.’ We add back interest to net profit because we want to calculate the profits available before the company distributes some to lenders. The reason why we exclude taxes, though, is a bit more complicated. Companies pay taxes on profits after they pay out interest, so profits before taxes give a better indication of the money available for lenders. But since we’re still trying to analyze companies from the shareholders’ perspective, there’s a good argument to incorporate taxes. Another consideration is that tax rates can be very volatile, with taxes owing sometimes coming in negative when the company receives a refund. Some analysts looked at the spotty data and said, “screw it, let’s just leave taxes out,” and that’s how EBIT, and not EBIAT (Earnings before interest after taxes), became the standard metric.

Historically, investing in low EV/EBIT ratio stocks would have outperformed the market more than investing in low P/E ratio stocks. The ratio derives its power from the same source as the P/E ratio – companies trading at low ratios are likely trading below their intrinsic value. Unfortunately, EV/EBIT shares many of P/E’s deficiencies as well. High EV/EBIT ratio stocks can still be cheap if the company holds significant assets that don’t affect profitability, or if the company’s profits grow substantially, or if accounting quirks underestimate the company’s true earnings potential.

In addition to these, EV/EBIT possesses one more deficiency that P/E doesn’t suffer from – it penalizes companies that can rightfully take on a lot of debt to run their business.

Regulated utilities are one such victim. Suppose a utility is pondering how to finance a $100 million power plant. They know with a high degree of certainty that the plant will generate $10 million in EBIT annually. If they finance 50% of the plant with debt at the cost of $4 million per year in interest, the shareholders will get (10 – 4) / (100 – 50) = 12% rate of return on their investment. But if the company finances 75% of the plant with debt at the cost of $6 million per year in interest, the shareholders will get (10 – 6) / (100 – 75) = 16% rate of return on their investment. Higher debt loads put more money in shareholders’ pockets, but the EV/EBIT ratio ignores this distinction between debtholders and shareholders.

One way out of this predicament is to use a variety of ratios in conjunction with EV/EBIT. For example, an investor can give some weight to each stock’s P/E or leverage ratio in addition to the EV/EBIT. Another method filters out companies using EV/EBIT before selecting stocks based on P/E or some other metric. However, these rules can become arbitrary, so I personally rely on machine learning to combine different ratios.

Despite its imperfections, the EV/EBIT ratio has become one of the most popular value metrics among quants because of its historical performance. In the next article, I’ll introduce another value metric that’s perhaps more popular among fundamental analysts than quants.