There’s a particular author on SeekingAlpha that is short Amazon (AMZN), and has been vocal about his rationale for the position; the only problem for him is that he’s been short the stock since early 2012 (as far as I can tell from disclosure on his articles), and the stock has nearly doubled over that time period. Every article he writes, no matter the validity of the argument presented, there’s bound to be a comment or two along these lines: “You’ve been short since the stock was at 200 and it keeps going higher – you’re wrong! Just give it up already.” That mindset isn’t confined to these few individuals – it seems to be common among many market participants. Some of it is undoubtedly due to the short-term nature of the paper shufflers, but there’s also a bit of disregard for one of Ben Graham’s most insightful statements:
Mr. Market is there to serve you, not to guide you.
Many investors (including myself at times) fall into the horrible practice of checking stock prices for answers. The reality is that it’s so simple to look at a stock you used to own, compare the current price to where you sold it at, and instantly conclude that you’re an idiot or a genius. The only problem is that stock price movements, even over a period of years, may not reflect the underlying changes in intrinsic value. Simply put, you must reevaluate the underlying business to draw a sound conclusion.
A great example of this is Cisco (CSCO) in the late 1990s and early 2000s. Let’s assume that you bought the stock in 1996 at $4 per share (split adjusted); two years later, the stock moved into the double digits. Looking at the fundamentals, you’re starting to get a bit concerned: With a forward P/E in the mid-30s and a market cap on the order of $70 billion, you’re wondering whether a company of such size will be able to sustain a level of growth that can justify the current price. With a two year CAGR above 50%, you decide to take your gains and call it a day.
For the next two and a half years (that’s approximately 650 individual trading days, with Cisco making new highs on plenty of them), you would’ve been hit with a constant barrage of price action telling you that you’re a complete idiot; before all was said and done, the stock would nearly crack $80 per share – 20x your original investment from less than five years ago, and nearly 8x the price when you left the party. Your last concern in this situation would be that the stock was now trading at more than 200x earnings, with a market cap of more than half a trillion dollars; all you know is you’ve made a horrible decision to sell all those years ago.
Fast forward to 2012: over the past decade, Cisco’s per share earnings increased from $0.25 to $1.86 – good for a stellar 10-year compounded annual growth rate of 22.2%; despite that, here’s what the stock has done over the same period:
If you had held the stock from the sell price of $10 to today (approximately $23 per share), your return over the next 15 years would’ve only been 5.7% per annum (before dividends); any purchases made after 1999 through most of the dot-com crash (February 2001) are still in the red. Of course choosing today’s price as our measure isn’t a perfect solution either – it’s just as arbitrary as choosing the peak in the early 2000s. What else can we do besides looking at the stock price?
My thinking is that you should remember your original rationale – in this case, selling a company with a $70 billion market cap because it would have a hard time growing at a rate to justify a forward price-to-earnings multiple approaching 40x. From here, I think you should be able to put rough numbers on what that growth might look like (after consideration for things like capital return to shareholders, the amount of financial leverage in the business, etc); maybe you decide that the only way the company can justify this current price would be to grow EPS 20% per annum for the next ten years (approximately half the rate of the five-year CAGR from 1994 – 1999; of course that was from a smaller base). At that point, assuming a more reasonable multiple of 15x on terminal earnings, the stock would trade at $27 per share ($1.80 x 15); after owning the shares for a decade, and watching earnings increase more than six-fold, the stock would’ve only increased 10% per annum. After consideration of the potential risks (namely not increasing earnings 20% per annum for a decade), this might seem grossly inadequate.
Even if you decided that was sufficient, you couldn’t have justified hanging around much longer as the stock continued to climb (if you stayed true to those assumptions). Shortly, you would’ve been taking on this outsized risk (I think it’s safe to say that a base assumption of 20% annualized earnings growth for a decade is far from conservative) for a mid-high single digit return; the math, which us highly dependent upon recognizing that a company of this size couldn’t grow at these rates forever – and would eventually have a multiple that reflected that – would’ve forced your hand. With hindsight, that is essentially what has happened – except the stock trades at 12x, not our “conservative” 15x.
You might quibble with my use of 20%, particularly after looking at earlier rates; what might cause one to take precaution in proceeding with the growth rate of the past few years, beyond the law of large numbers? Here's an important disclosure in the company’s 10-K, from the late 1990’s:
“We expect that in the future, our net sales may grow at a slower rate than experienced in previous periods, and that on a quarter-to-quarter basis, our growth in net sales may be significantly lower than our historical quarterly growth rate. As a consequence, operating results for a particular quarter are extremely difficult to predict.”
Note that management didn’t say that sales growth could slow in the future; they said “we expect” net sales growth may slow (in earlier years they had been even more blunt, stating "net sales will grow at a slower rate than was experienced in previous periods"). There’s a subtle but meaningful distinction between the two – especially for a company with the built-in expectations that CSCO had. The company also stated that it expected pressure on both gross and operating margins in the future; this puts even more pressure on top-line growth, starting from a sizable base of $6 billion in fiscal year 1997. Matching our previously stated growth rate of 20% per annum over the coming decade would bring that figure to $37 billion in 2007. Think about that for a second: This would require Cisco to add an incremental $31 billion to the top line in just 10 years. By way of comparison, it took Procter & Gamble (PG) more than 150 years to reach $31 billion in sales.
This discussion warrants one additional short story: in the late 1980’s, Peter Cundill was convinced that the Nikkei in Japan was grossly overvalued; for three straight years, he continued to hold put options against the index, which were a perpetual headwind to the performance of the Cundill Value Fund. In Christopher Risso-Gill’s fantastic book, “There’s Always Something to Do,” he recalls a journal post from Mr. Cundill years into the (at the time) unsuccessful trade:
“Last night I was anticipating the outcome of the Japanese stock market. I dreamt that the Nikkei fell by a thousand points. In fact it was up 1,450 – the second best in its history. You must stop this short term anticipation stuff. If you’ve done the numbers and are satisfied with them and the principle is right, you just have to grit your teeth and be patient.”
In an interview a few years later, even Cundill, with a track record that puts him amongst the best of all time (more than 15% per annum for three decades), admits that he almost gave in and let the continued negative information from the market push his hand:
“I almost stopped selling Japan short in the last quarter of 1989 because I couldn’t stand it anymore. But intellectually I was convinced I was right and so I carried on – and then in the first quarter of 1990 the Japanese market fell 25% in eight weeks and I made back everything I’d given away since 1987, plus a good deal more.”
Here’s the point: You can’t look at Cisco at $20, $30, $40, and (eventually) $80 and conclude that the price action alone proves that you’ve made an investment error; the fundamentals are just as important in retrospect as they were in the original decision – they cannot be overlooked.
The merit of an investment decision cannot be decided in hindsight by studying near term price performance – and the near term can at times extend many years into the future (Isaac Newton famously let the price action drive his activity during the South Sea Bubble – and turned a nice gain into a substantial loss of more than 3x his prior profits). The only way to reassess the validity of your original decisions is to compare the company’s financial performance with reasonable ex-ante assumptions. You’ve got to decide if the market is there to serve or guide you – and then act according to that belief. Reflection upon past investment decisions must be looked at in the context of business performance (while also remembering what was unknowable at the time), rather than the performance of the stock on its own; thinking about previous buy or sell decisions in this context should lead to improved analysis and decision making over time.
That’s a long way of saying that laziness cannot replace analysis – and even though everybody here understands that, the simplicity of doing otherwise is often irresistible; the next time you find yourself looking at market quotes for answers, it might be worthwhile to think twice.
About the author:
I think Charlie Munger has the right idea: "Patience followed by pretty aggressive conduct."
I run a fairly concentrated portfolio, with a handful of positions accounting for the majority of the total. From the perspective of a businessman, I believe this is sufficient diversification.