To Sell or Not to Sell

If the job has been correctly done when a common stock is purchased, the time to sell it is – almost never.

Philip Fisher

Have you heard? Apparently, stocks only go up. Curious narrative given what transpired a mere four months ago. Retail traders have been piling in at alarming rates, particularly into highly priced, story stocks and/or financially distressed companies. Anecdotally, I was in a conversation when an aspiring actress said “I have no idea what investing is, but I was given a free stock if I opened a Robinhood account, so now I have one share of ADP! I don’t even know what ADP is!” This story is certainly not a one-off. Increased financial participation is great in the context of a well thought-out financial plan, but putting money at risk in the stock market can lead to financial ruin, particularly for those who don’t know what they’re doing, and may even have dreadful, tragic consequences. The phrase “history doesn’t repeat itself, but it often rhymes” comes to mind when observing such exuberance on behalf of those less experienced. But when others are gambling, there is no need to participate. Going one step further, should it warrant caution?

In general, it pays to be contrarian at extremes. In March, when prices were depressed due to completely valid concerns over the looming economic impact of the coronavirus, opportunistic buying worked out brilliantly given the herculean rebound in prices. Was it skill? Luck? Both? Whatever the case, the backdrop has now shifted; the economy is far from recovered with many states and businesses struggling to re-open, yet major indices are back to levels reached before the crisis onset. Much to the bewilderment of those who attempt to explain market moves, “disconnected from reality” has become a widespread narrative. For those who do not, like me, it’s nevertheless difficult not to sympathize to a certain degree with such a statement given the opaque visibility for businesses that must operate in the real economy. Obviously, some businesses have been, and continue to be, far more impacted than others and the market has rightfully attempted to discount those impacts accordingly. Not every stock has rebounded with the markets; the bifurcation between the haves and have nots is readily apparent. Yet at the same time, there are certainly have nots that are being priced as having. At a time when a myriad of stocks are priced very highly compared to their underlying fundamentals, such as revenues, earnings or cash flows, is it time to sell?

Far more words have been written about what, when, and why to buy than on what, when, and why to sell. Some of the great investors, though, have weighed in. Howard Marks believes defensiveness and caution are warranted when exuberance and optimism are evidently ubiquitous, indicating holding less risky, higher rated assets with more downside protection is prudent. Peter Lynch had rules for selling depending on the category of stock held – such as if a slow grower’s price has appreciated despite deteriorating fundamentals, if a fast grower’s phase of rapid growth is ending, or if a turnaround has turned around – which make sense because different businesses have different characteristics and different catalysts for unlocking value. Joel Greenblatt concurs, quipping, “trade the bad ones, invest in the good ones,” meaning an average company in a competitive industry bought at a bargain price should be sold after a notable rise in price whereas a good business should be held. Benjamin Graham’s strategy was more of the former, buying many mediocre businesses extremely cheaply and selling “at such times as they are dear, or at least no longer cheap, by analysis” or if they “can be replaced by issues much more reasonably priced.” Philip Fisher, who was at the opposite end of the spectrum in buying a group of outstanding businesses with high growth prospects for the long run, listed three reasons for selling: when the investor has made a mistake in that the factual background of the business is significantly less favorable than believed at purchase, changes due to the passage of time that cause a business to no longer qualify for his reasons for purchase, and when another opportunity presents seemingly materially better prospects than a current holding.

It is well known Warren Buffett’s favorite holding period is “forever,” maintaining, “the best thing to do is buy a stock you never want to sell” and “the real thing to do with a great business is hang on for dear life.” Charlie Munger, in typical fashion, agrees that, “if you buy a few great companies, then you can sit on your ass. That’s a good thing.” Although “forever” is their favorite holding period, it isn’t their only holding period. Warren has said, “we sell really when we think we’re reevaluating the economic characteristics of the business… [when] we don’t think their competitive advantage is as strong as we thought it was when we initially made the decision” and Charlie that, “the sales that do happen, the ideal way, is when you’ve found something you like immensely better.”

Sell decisions are notoriously far more difficult than buy decisions. Capturing the complexities of every situation is impossible, but some themes emerge. Considerations for selling will depend on Mr. Market’s mood, the type of holding, the reasons for the initial purchase, analysis of the underlying business trends, and other potential opportunities. Synthesizing these points further, below is my own simple, overarching methodology for determining when to sell. Importantly, it will only be relevant for those who adhere to a similar investment style of buying equity ownership in great businesses with defensible competitive advantages and opportunities for adept management to deploy capital at high rates of return over the long-run, and thus won’t be relevant for every investor.

First off, the following are NOT good reasons in-and-of-themselves to sell a position:

• The price has gone up a large percentage. Rebuttal: There is no arbitrary rule that the price cannot go up more.
• The price has reached a price target. Rebuttal: Price targets convey implausible precision and will be obsolete without consideration for what has transpired in the interim.
• The price is hitting “resistance.” Rebuttal: Anybody following technical analysis is more likely a speculator than investor.
• To lock in gains. Rebuttal: Realizing unrealized gains may be desirable, yet what is undesirable is missing out on further gains.
• The position has become large percentage of the portfolio. Rebuttal: Unless explicitly prohibited from or burdened by holding a position above a certain percentage, there is no external justification to trim a winning position simply because it has become large. Further, a position will become outsized if it either has a large cost basis, suggesting conviction in the idea, or if other positions have done poorly, suggesting those deserve reconsideration beforehand.
• The stock is loved and over-owned. Rebuttal: It’s impossible to tell if positioning is overcrowded, no matter how many funds hold it or how high the percentage of surveyed investors claim it so.
• The multiple is higher than historically. Rebuttal: Businesses can transform to become better than they were previously and/or the market may realize the business is better than previously believed, in which case an increase in multiple is likely justified.
• The market is highly valued. Rebuttal: Conflating the micro and macro or trying to time the market are probably the most cited reasons for selling, making them also the most common mistakes. Not only is it impossible to consistently time the market, it’s impossible to consistently time the market. Phil Fisher more eloquently explains, “If the argument is valid that the purchase of attractive common stocks should not be unduly influenced by fear of ordinary bear markets, the argument against selling outstanding stocks because of these fears is even more impressive.”

The above are, however, possible indications of what are, in my view, the only two valid reasons for selling:
1. When the price is considerably higher than the intrinsic value.
2. There is another more attractive opportunity to allocate capital to instead.

When buying, there is an implicit assumption that the market is wrong, that the intrinsic value is higher than the current price, and that the price should converge to it over time. Why, then, can’t the market also be wrong in pricing a stock higher than it’s intrinsic value? If an investor assumes price and intrinsic value are independent, it seems logical they can diverge in either direction. Again, intrinsic value is just a subjective estimation of the worth of a business based on its discounted cash flows, growth, and risk. The estimation should have a confidence interval given the predictability of the business and the irreducible unknowns and uncertainties. It’s impossible to precisely forecast the cash flows a business will generate over its lifetime, but it’s still worthwhile to have a general sense of the realistic possibilities and the risk and rate of return implied by various prices. Thus, incorporating the important concept of a Margin of Safety provides the investor some leeway for being wrong and not losing, or gaining, much.

Building on that concept, it seems intellectually consistent and internally justifiable to apply the same methodology to both buy and sell decisions. Both require a determination as to whether the price provides adequate reward for bearing risk and uncertainty, and then acting on your convictions. A Margin of Safety on the upside is useful because an investor could also be wrong in sell decisions!

Understanding why the price has appreciated is therefore integral to making a sell decision. Have the fundamentals or their outlook improved, validating a rise in price? Or has price risen solely because the entire market has increased, with little regard to underlying static or deteriorating fundamentals? Especially during short-term disruptions, such as the covid pandemic, some companies may temporarily benefit or be hurt, but will likely revert back to normal over time. Others, though, will experience prolonged effects. Evaluating the lasting implications beyond the transient impacts will prove to be a much more valuable use of time than vice versa.

Improvements in or erosion of a company’s competitive advantage or ability to endure are vastly more influential than a bump or drop in sales or earnings due to cyclical factors. Any changes that affect the business’s ability to generate excess returns on capital, such as changing industry dynamics, an evolving competitive landscape, indications of pricing power, or reinvestment opportunities, should be scrutinized intensely. Guesstimates around next quarter’s numbers, not so much. Some investors like to chase the short term trends, piling in to the obvious winners. But mean reversion can be ruthless.

Unlike the Federal Reserve, investors deal with a limited capital base so their decisions must also incorporate alternative opportunities that may have higher expected risk-adjusted returns. In the context of one’s portfolio, there is some credence in re-balancing; selling some of what has performed well to buy more of what has not. This works to a lesser degree for individual stocks than it does for asset classes, yet it’s possible to enhance portfolio returns versus a simple buy and hold strategy. Think about it this way: since stocks that increase in price become a larger part of the portfolio at the same time as their forward return expectations are reduced, the portfolio’s expected return will be lower compared to re-allocating into the stocks with lower weights and higher expected returns. This, of course, is far from rudimentary or a free lunch; it’s dependent on the investor’s ability to accurately gauge the intrinsic values of all held stocks at all times and the costs of transacting. It’s certainly possible that a price increase equals or even lags the increase in intrinsic value, in which case the expected return is still reasonable and should not be sold.

The biggest argument against selling is one of the two unavoidables. Not death, but taxes. Evidence has shown that low turnover portfolios outperform high turnover ones precisely for this reason. The drag from capital gains taxes weighs heavily on compound returns. Too often, investors’ succumb to short-term time horizons, leading to more frequent turnover and a large chunk of gains being wired to the tax man rather than pocketed. Investing with a long-term time horizon means sell decisions should come few and far between, but being patient amidst the noise isn’t easy.

Besides the tax implications, behavioral biases also weigh on good decision making. A recent study found that managers actually do generate alpha with their buy decisions, but detract from it with their sell decisions. This is because they “use heuristics when selling but not when buying.” These “costly, systematic biases” are “consistent with PMs devoting more cognitive resources to buying than selling” producing underperformance “explained by a heuristic two-stage selling process, where PMs limit their consideration set to assets with salient characteristics (extreme prior returns) and sell those they have the least conviction in (low active share assets).” Stocks that have risen precipitously produce large unrealized gains, tempting investors to realize those gains, before they potentially disappear.

I suggest that instead, flip the analysis around and evaluate the potential further gains that would be forfeited by selling too early. Notably, if it’s a great business, the future gains could be multiples of the current ones. Additionally, price appreciation should be expected of held positions, or else why were they bought in the first place? Peter Lynch famously said, “Selling your winners and holding your losers is like cutting the flowers and watering the weeds.” Another common mistake is anchoring to past prices, even though news evolves rendering previous prices and forecasts obsolete. The only “price” you should anchor to is your estimate of intrinsic value, which should itself fluctuate, though less wildly, as new information unfolds.

Inevitably, hindsight bias will affect the assessment of the decision based on the outcome. As sure as the sun rising, price action will cause angst because “of course you shouldn’t have sold” if the price rises or “of course you should’ve sold” if the price declines. But the actual outcome was only one of many possible outcomes. Ignore the subsequent price action and focus on the decision process. A good practice is to write down the basis for every buy and sell decision in order to mitigate hindsight and outcome bias. Documentation of what you thought at the time of the decision and what got right and wrong is beneficial for distinguishing the relative influences of skill and luck. No decision-outcome will be black or white, but being able to review and analyze your thought process will greatly improve the quality of your decision-making going forward.

Selling a great business simply because the price has risen or it appears optically expensive is often a mistake. Great businesses have the ability to continually defend and expand their competitive moats for decades, rewarding patient investors handsomely. If the company has a long runway for value-added growth and a competent management team at the helm, it should be reflected in an ever increasing intrinsic value, preventing the investor from mistakenly selling.

Although rare, great businesses with strong competitive moats can become overpriced and trade sideways for decades. No one can or should try to time the market, but price matters to expected and realized returns. When multiples are high, expectations and optimism are also high, likely signaling lower expected returns. If price shoots beyond the intrinsic value to a noticeable degree, selling and allocating capital elsewhere is justifiable, even when capital gains taxes must be paid. Nothing precludes a stock from being bought again, even if the price goes nowhere for a while. Though trying to dance in and out based on market sentiment is a fool’s errand.

Arguably, Buffett should’ve sold Coke when it was priced at more than 45x earnings. On the other hand, Coke was an amazing company with an intact secular growth story so why fret over a temporary overvaluation when selling would generate a massive tax bill, unnecessary attention, and misinterpreted signaling? He was definitely not looking to buy more at those prices, yet did advocate for Coke buying back shares, which was probably not the best capital allocation decision since the stock did not eclipse it’s all time high until more than a decade later.

Even the greatest can perhaps err when it comes to sell decisions. Because they’re hard, intellectually and emotionally. But when applying the exact same process of comparing price to intrinsic value for both buying and selling, and accepting you’ll make mistakes in each, hopefully it becomes a little easier.

To end, an excerpt from Seth Klarman:

Many investors are able to spot a bargain but have a harder time knowing when to sell. One reason is the difficulty of knowing precisely what an investment is worth. An investor buys with a range of value in mind at a price that provides a considerable margin of safety. As the market price appreciates, however, that safety margin decreases; the potential return diminishes and the downside risk increases. Not knowing the exact value of the investment, it is understandable that an investor cannot be as confident in the sell decision as he or she was in the purchase decision.
To deal with the difficulty of knowing when to sell, some investors create rules for selling based on specific price-to-book value or price-to-earnings multiples. Others have rules based on percentage gain thresholds; once they have made X percent, they sell. Still others set sale price targets at the time of purchase, as if nothing that took place in the interim could influence the decision to sell. None of these rules makes good sense. Indeed, there is only one valid rule for selling: all investments are for sale at the right price.
Decisions to sell, like decisions to buy, must be based upon underlying business value. Exactly when to sell – or buy – depends on the alternative opportunities that are available. Should you hold for partial or complete value realization, for example? It would be foolish to hold out for an extra fraction of a point of gain in a stock selling just below underlying value when the market offers many bargains. By contrast, you would not want to sell a stock at a gain (and pay taxes on it) if it were still significantly undervalued and if there were no better bargains available…
If selling still seems difficult for investors who follow a value-investment philosophy, I offer the following rhetorical questions: If you haven’t bought based upon underlying value, how do you decide when to sell? If you are speculating in securities trading above underlying value, when do you take a profit or cut your losses? Do you have any guide other than “how they are acting,” which is really no guide at all?”