Six times earnings, a dividend above most bonds, a recovery thesis that fits in three sentences. It is worth being precise about which year those figures describe.
Consider a stock at about six times trailing earnings. The dividend yields more than most bonds pay. There is debt on the balance sheet, though nothing that would keep anybody awake, and the industry is thoroughly out of favour, which is precisely where the books tell you to look. The recovery thesis fits into three sentences.
So an investor buys. The price falls. He buys again, five times earnings being cheaper still than six, and feels rather clever about it... which is a sentence that has never once ended well. What happens after that is worth following closely, because at no stage does the analysis break down on its own terms, and the position still ends where it ends.
There is a particular silence that follows the sentence “so I bought some more of it”. I have heard it across our own kitchen table, from a wife who could strip paint with it! It outlasts the explanation by some margin, particularly when the explanation involves the phrase “averaging down,” and the listener has already worked out what that means without needing the technical term for it, which is usually a fair sign that the technical term was doing work it had no business doing. She had.
What the multiple was measuring
A P/E of six is today’s price divided by last year’s earnings. That is the whole of it. The figure feels like a statement about value. It is a statement about the past.
Suppose those earnings are about to halve. The real forward multiple is twelve, an ordinary number for an ordinary business, and the stock was never cheap at any point in the story. The screen measured the wrong year. So did the investor.
Anybody who has read a book on value investing knows this in the abstract. The harder thing to absorb is that a low multiple carries a forecast inside it, the market’s own forecast, and buying at that multiple places a bet against it. The investor is telling a large number of well-informed people that they have got it wrong, which is a perfectly respectable thing to do once the work has been done, and in the case above the work has not been done.
The sign in plain view
Most value traps share one uncomfortable feature. The warning sat in the accounts the entire time.
Revenue had been falling. That much the investor had seen and filed under cyclical. Gross margin had been falling too, three years running, and this was noticed and then passed over as temporary pricing pressure that would reverse when the cycle did, an explanation with the convenient property of requiring nothing whatever to be done about it.
The distinction here is worth more than most of what gets written about valuation. Falling revenue alongside steady margins is a volume problem, and volume problems recover. Falling revenue alongside falling margins means the company is cutting price to keep customers from walking, which means its pricing power is going, and pricing power does not come back with the cycle. The business is becoming a worse business. The value traps guide puts that distinction first for exactly this reason.
Why the buying continues
Each drop in the price took the multiple down with it, and the paper case for adding improved every time. Averaging down has no stopping point. The average improves at every lower price, so the method invites more buying at precisely the moment the evidence is turning against you.
In the illustration, three additions carry the position to more than twice the size it was meant to reach, in a business the investor has a thesis about and no real edge in. Then the dividend is cut. The payout ratio had been predicting that quietly for some time. The price falls hard on the announcement and the position is finally sold.
The loss itself is survivable. The larger cost is a couple of years of capital sitting in something that went nowhere, while better-researched ideas moved along without it, and that sort of cost appears on no statement anywhere, gets included in no performance figure, and is for that reason the part of the episode almost nobody troubles to count.
Three rules that follow
All three are mechanical, on the grounds that judgement in the middle of a falling position cannot be trusted, mine included.
Check the margin trend before you look at the multiple. Gross margin down for more than two consecutive years keeps a stock off the shortlist, however cheap it appears on a screen. Cheap with eroding margins is the specific shape a value trap takes.
Set the maximum position before the first purchase. Treat it as a hard ceiling. Add on the way down if you must, up to that ceiling. Then you stop. The stock average calculator tracks the blended cost, and it does something more useful besides, because it shows how large the position has quietly become while you were thinking about the average.
Write the fundamental exit before entering, as a condition and not as a price, since a multi-year position will pass through a great many prices. Something on these lines: “if gross margin falls again next year, I am wrong, and I sell.” A condition of that kind survives volatility and still fires when the business deteriorates.
For anyone starting
Cheap is a forecast wearing the clothes of a fact. Before buying on a low multiple, write down what the market has to be wrong about for the trade to work, and then ask whether you know anything in particular about that specific question. An honest “no, but it looks cheap” means you are taking the other side of a bet against people who have done more work than you have.
The key stock metrics guide repays reading with one question held in mind: which of these numbers describe last year, and which describe next year? Most of the ones that make a stock look cheap describe last year.
Comments
Log in or sign up to join the conversation.