Monday, May 23, 2016

Predicting long-term winners is difficult

In the past few years, compounders—companies that can steadily compound their earnings and equity value over long periods of time—have captured investors' attention.

The allure of compounders is easy to see: Day traders have to find new trades all the time. Event-driven investing is less hectic than day-trading, but event-driven investors still have to search regularly for new ideas. By contrast, buying compounders necessitates no endless search for new investments. Instead the compounders naturally grow, doing the investor's work for him.

As elegant as that is in theory, it's not always easy in practice. Some companies that look like long-term winners turn out to be dogs. Valeant and Ocwen Financial are two faux-compounders that have imploded recently. The opposite can also happen: some apparent dogs turn into long-term winners, with their success being largely unforeseeable. Here are three examples.

Atlantic Tele-Network

Atlantic Tele-Network (ATNI) acquired Guyana's telephone monopoly in 1990. During the 1990s, it acquired other telecom assets in the Caribbean before splitting them off into a separate public company in 1998. Since then, ATNI's stock has appreciated 2150%—a 19% annual return over 17 1/2 years. Dividends increase the annualized rate of return to 24%.

Two acquisitions are responsible for most of this prodigious appreciation. In 2005, ATNI acquired Commnet, which operates "wireless base stations" in sparsely populated areas. It paid $59 million for Commnet, which will produce $75 million of EBITDA this year.

In 2009, ATNI acquired 900,000 wireless subscribers from Alltel when it merged with Verizon. ATNI paid $200 million for the subscribers, invested another $100 million in the acquisition, and sold it to AT&T for $800 million in 2013.

But before 2005, there was nothing in ATNI's history that suggested it was capable of making successful acquisitions. From 1990-98, ATNI's stock lost 60% of its value. The Caribbean telecom company that was split off in 1998 later went bankrupt. In 1998 and 2000, ATNI bought two Haitian telecom companies that it had to write off in 2001. In 2001, it invested in a bandwidth-trading company that filed for Chapter 7 bankruptcy less than a year later. Other money pits included a Guyana-based call center and a collect-call business. The entrepreneurial acquisition strategy that eventually made ATNI a home run incinerated money for years.

Tractor Supply

Tractor Supply (TSCO) operates "rural lifestyle retail stores," i.e. stores that cater to hobby farmers. Since its all-time low in late 2000, TSCO stock has appreciated a phenomenal 18,700%, or 40% annualized over 15 1/2 years. The company has achieved these returns by retaining most of its earnings and reinvesting them at high incremental returns.

But before 2001, TSCO's results suggested it was a below-average business. From 1995-2000, TSCO grew earnings per share only 6% per year despite retaining all earnings. It earned $70 million from 1995 through 1999, yet earnings were only $4.25 million higher in 2000 than in 1995. From 1994 to 2000, TSCO stock fell 60%.

In early 2001, soon after the stock began its upward rampage, "Charlie479" from Value Investors Club wondered if TSCO's business model doomed it to poor returns:

It seems to me that it's competitive strategy is precisely the one that anchors it to mediocre returns on capital. If the stores' attraction (indeed, the key to their 5-10 year survival against the big boxes) is the hard-to-find, slow-turning item that the Home Depots or Lowes won't stock, then a large inventory is a permanent part of this company's operating model. This large working capital requirement translates into high capital requirements for the business... I think the economics of the business will produce only average market returns over the long term.

Ball Corporation

Ball Corporation (BLL) makes aluminum cans. Since March 2000, BLL stock has risen more than 2000%, or 22% annually. Over time, the can-making industry has gradually consolidated into an oligopoly, increasing the pricing power that BLL and its competitors enjoy. Notably, BLL has been active in acquiring its rivals, so it's been both an instigator and a beneficiary of this trend.

Consolidation's benefits didn't materialize immediately, however. BLL and its competitor Crown, Cork & Seal made numerous acquisitions during the 1990s, yet in early 2000 BLL's stock traded at a price it had first reached in 1985. Even including dividends, the stock barely kept pace with inflation over that fifteen-year period.

Another thing these three companies have in common, along with the improbability of their stock-market performance, is that they've all benefited from multiple expansion. At their 2000 lows, BLL and TSCO traded at 5.5x trailing free cash flow and 4.5x trailing earnings, respectively. Today they trade at 20x and 27x this year's estimated earnings.

ATNI traded at less than 3x earnings in 1998. Today it's difficult to value because it owns a mix of businesses that produce steady earnings and asset-rich projects with minimal reported earnings, but one investing pitch suggests that the market is valuing ATNI's operating businesses at double-digit earnings multiples.


  1. This is a very nice article YM! Unfortunately, the amount and timing of future total returns is unpredictable. A total return could be 20% over 30 years, but this could include 15 years of no returns, followed by 15 years of high returns. The first 15 year would have tested the patience of even the most patient investors out there. Most would have likely sold out very quickly, because "the investment was not working". If you are also benchmarked against something like S&P 500 which did better in the first 10 -15 years, and the chances of selling out would have been even higher.

  2. Young Money: Once again great post. My only quibble is why so infrequent of posts. This needs to change immediately.

  3. It's a great post. I've thought about two general categories of investment: one category with businesses that are doing well now but are priced as if they won't be doing as well in the future. And businesses that look expensive now but could actually be extremely cheap if they are able to execute on their plans for future growth. (The stocks that have unforeseen positive developments occur aren't really relevant, as they are usually impossible to predict). The first category usually consists of the larger companies that get sold off in market corrections or maybe due to some sort of temporary or short-term issue. There are many more of these opportunities. The second category are the ones like TSCO; the dream investments that compound at high rates. They are out there, but very difficult to predict, as you point out. I do think it's valuable to look for companies that are able to retain earnings and reinvest at high rates, even if there is uncertainty whether that will continue (I don't know of any situation that is certain anyhow?), as those businesses sometimes offer upside if things go well, but not as much downside if the great futures don't end up developing...

    Just some random thoughts as I read the post. Great work, and I completely agree with Matt on the posting frequency.

  4. Thank you for your work on these. If we agree that predicting is nearly impossible, should we then bother spending any time or effort in trying to make compounding returns achievable?

    To achieve wildly fantastic compound returns in such names requires one to stumble into them (by whatever mental mechanisms you like), but more importantly, to resist every urge to sell the three-baggers. Even if I believed I had great foresight into these businesses, and was resolute enough to commit significant capital to them, I would have to forego any temptation to book my well-deserved gains.

    It calls to mind Ben Graham's Postscript to the Intelligent Investor, in which he seemingly admits that his most optimistic predictions of Geico's business would have never reached the mark that the company ultimately achieved. And worse, the market price always seemed to advance faster than Geico's growth in profits.

    The only reason he achieved the immense returns that he did, was because he regarded Geico as a sort of 'family business', and consequently, he "continued to maintain a substantial ownership of the shares despite the spectacular price rise."

    Graham concludes: "Ironically enough, the aggregate of profits accruing from this single investment-decision far exceeded the sum of all the others realized through 20 years of wide-ranging operations in the partners' specialized fields, involving much investigation, endless pondering, and countless individual decisions."

    For me, the most critical element in the Geico success and in Graham's investment career was his sense of feeling bound to a successful and growing company, coupled with his willingness to accept its high market valuation.

  5. YM,

    How will this information influence(change) your approach to investing?

  6. Thanks for all the comments!

    "Unfortunately, the amount and timing of future total returns is unpredictable..."

    Yes, that's a good point. I think your comment describes what happened to most value investors in the late 90s: they owned businesses that ended up producing great L-T returns, but they underperformed for long enough during the tech bubble that they lost most of their clients.

    " I do think it's valuable to look for companies that are able to retain earnings and reinvest at high rates"

    Agree, and I enjoyed the guest post at Base Hit Investing that talked about this. One practical problem is that some of the companies with the best growth/reinvestment opportunities are big stock-options issuers. Which isn't necessarily a deal-breaker, but it eats into the appreciation potential.

    "Thank you for your work on these. If we agree that predicting is nearly impossible, should we then bother spending any time or effort in trying to make compounding returns achievable?"

    I'll resist the urge to be overly cynical and say it's impossible... In some cases, I think it's is at least semi-predictable. In 2000, Ball's management was buying back a lot of stock at low valuations and, while industry consolidation wasn't yielding huge benefits yet, it certainly wasn't hurting the company either. I think someone who looked at the company back then could have said, without the benefit of hindsight, that it was a decent bet and the company probably had its best days ahead.

    "For me, the most critical element in the Geico success and in Graham's investment career was his sense of feeling bound to a successful and growing company, coupled with his willingness to accept its high market valuation."

    That's an interesting idea, i.e. that Graham stock with his greatest investment for (in part) sentimental reasons. Anything that helps one stick with a winning investment is good, even if it isn't purely rational.

    Going by memory here, but I read once that GEICO fell 95% from the all-time high in the last few years of Graham's life. I hope he was out by then!

    "How will this information influence(change) your approach to investing?"

    I'm more of a trader than a buy-and-hold investor, so it mostly confirms what I already think about the difficulty of forecasting very far out. It also gives me a sense that business quality isn't immutable, and businesses can be better (or worse) than they look, so it's a bad idea to put companies in rigid boxes and label them inherently low-quality or high-quality.

  7. The TSCO case in particular shows that compounders investing is very hard, considering that "charlie479" is no other person but Norbert Lou of Punchcard Capital, who is famous for earning high returns by running a concentrated portfolio of compounders.

    I think most individual investors has no business with compounders investing since their capital is so small they can capitalize on other pockets of value on the market. To me, compounders investing belongs to the too-hard pile. I will stick with special situations with a strict adherence to Ben Graham's teachings (chapter 8 and 20).

    Hong Kong


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