Monday, October 6, 2014

Avoiding value traps

Although I panned Daniel Strachman's book about Julian Robertson in my last post, the book has one great quote that's worth sharing:

Robertson’s mantra was, as long as the story around the investment remained the same, the position should get bigger. As soon as the story changed, it was time to get out...

To understand the concept of story, consider this example. Say you are interested in a solid oak wooden table. [An] analyst could tell you that he had checked out the market for tables, evaluated the information, and come to the conclusion that the table was a good buy at $100 because it was well made, solidly built, and would not fall apart. This is the story. So you go to the shop, prepared to buy the table. And then, just as you are running your hand over the table, a corner falls off. Well, now the seller is desperate to get rid of the broken table and is willing to sell it for $20. To the analyst, this seems like a steal. He sees an incredible opportunity to buy something for $20 that is really worth $100 and needs just a bit of fixing to get it there. But in Robertson’s eyes, the story is now flawed, and now he would say that you should want no part of the deal. How could something so well built, made of the finest oak, break?

This is a simple analogy, but it conveys a profound truth about value traps and how to avoid them.

When a company with a long record of success starts performing badly, investors tend to view the present in light of the past and assume its problems are temporary. If the stock goes down, they'll see it as a buying opportunity-- witness the numerous VIC writeups on Dell, Staples, and Tesco. Instead, investors should ask if the bad performance indicates that the company's competitive position has permanently deteriorated. If it has, then the company's historical performance and the fact that its stock is cheaper than it used to be are meaningless.

The same principle applies to valuation standards. One of my first posts on Young Money was about gold miners. Although they were trading far below their historical valuations, I argued that they were value traps: since the price of gold had been low for years, investors historically had ignore the miners' low earnings and valued them on the profits they would earn if gold surged. But when I wrote that post, gold had already surged and the miners' earnings were still low. The assumptions that investors had historically used to value the miners had been disproven, so it didn't matter where they traded relative to historical levels.

I used to be a hardcore statistical value investor, and it took me years of experience to realize how important this is. If a business is declining or its competitive environment is changing in a way that will permanently impair its profitability, there's almost no price at which it's worth buying.

The problem goes beyond falling earnings. When a company's business model deteriorates, management typically doesn't know how to respond. If they're used to running a growing company, they won't know how to run a shrinking company. If they're used to dominating their industry, they won't know how to compete from a position of weakness. In many cases, they'll do things that exacerbate the company's problems.

Their capital allocation also typically worsens: it's common for structurally-declining businesses to try to grow their way out of their problems by making overpriced acquisitions in growth markets. See HP and Autonomy, Kodak, Post Holdings, etc. Other structurally-challenged businesses try to maintain their earnings per share by buying back stock. This often starves their operations of necessary cash and can actually lower EPS in the long run.

For a company with a durable business model and long-lived assets, bad capital allocation isn't necessarily fatal. If the company wastes the next few years' earnings on overpriced acquisitions, that impairs only a small part of its present value. By contrast, the next few years represent a huge share of present value for companies that are in decline or that face a step down in profitability because of new competitive pressures. It's a sad irony that the companies for which capital allocation is most important tend to have the worst capital allocation.

As an example of this, Best Buy, Gamestop, and Western Union are all buying back lots of stock. These business face existential threats and could go out of business within the next decade. In my opinion, a company like that should never buy back stock-- it should try to dividend as much of its cash to shareholders as possible before it goes to business heaven. I feel the same way about retailers and tech companies-- most of them operate in an environment of constant change and have no residual asset value, so the surest way for them to reward shareholders is to pay dividends or special dividends while business is good.


  1. Good article! I don't think that Western Union will go out of business though. They are currently struggling due to bad management but technology will not disrupt their business. Their core business is in serving unbanked customers. Not everybody in the world has a bank account, even in the USA and even if they are in the USA legally.

  2. Well-said and I agree. It's worth pointing out that when the indecisiveness and poor allocation decisions are induced by the fear of the unknown, it is also likely that CEOs resort to the voice of the market that is filled with myopic thoughts -- thus the unfortunate decisions. Furthermore, now that activists wield a powerful weapon, a genuine attempt for change could be treated as travesty and gets nipped in the bud by public campaigns. A rather unfortunate predicament for public companies entering sunset indeed.

  3. Thanks for the comments!

    Regarding WU, I wouldn't dismiss the technological risk. Many poor countries have high cell-phone penetration rates, e.g. Philippines, Russia, and even El Salvador are over 100%. I think it's inevitable that they'll have high smartphone adoption once the phones get cheap enough, and these will eventually form the basis of rival payment networks. This article paints an interesting picture of how it might work:

    That company doesn't seem to have succeeded, but I think the idea is more likely to gain traction now vs. 2007. The big question- which the article mentions- is what kind of regulatory hurdles rivals systems will face, and I don't know the answer to that.

    Since WU has negative working capital, it will face severe problems if/when new technologies cause volumes to decline... I think of the working capital deficit as a contingent liability that comes due at the worst possible time.
    In the meantime, WU used to have 3x as many agents as Moneygram and now it's closer to 50% more, and Moneygram is used to earning lower margins.

    1. In payments the problem isn't really technology. It's about dealing with fraud and other operational issues.

      Paypal is interesting because they originally wanted to do payments from palm pilot to palm pilot. Very different than what they do today. They do have a lot of technology/algorithms to automatically detect fraudulent transactions. It's paypal ability to deal with fraud that allowed it to succeed for high-risk online transactions... while large numbers of its competitors failed and exited.

    2. Correct me if I'm wrong, but Paypal and Western Union have different business models. Paypal facilitates commercial transactions that are funded electronically. WU facilitates remittances from the sender to his or her relatives that are funded with cash. Fraud seems like a much bigger risk for the former than the latter. (I know WU offers online transfers and can be used for commercial transactions as well, but cash remittances are the vast majority of transfers.)

      For cash-funded transfers, their anti-fraud provisions are pretty basic. It's possible to receive money through WU without presenting an ID. In electronic transfers where the sender can defraud WU, I'm not sure what their technology is like vs. that of Xoom etc. This article suggests that Xoom's fraud losses are a small fraction (~20%) of losses in online retailing and low absolutely:


Note: Only a member of this blog may post a comment.