Tesla - Driven By Noise, But High-Quality Noise

| About: Tesla, Inc. (TSLA)
This article is now exclusive for PRO subscribers.


The fact that TSLA stock can’t be justified on the basis of value is not a deterrent from owning the shares.

Market equilibrium does not require price to equal value; price actually equals value plus noise.

Among large-cap stocks, we can make money through noise just as readily as we can through value.

The key is to differentiate between high-quality noise and low-quality noise.

TSLA can be justified on the basis of high-quality noise; the key is to make sure you take the trouble to do the grunt work.

Let's talk about Tesla Motors (NASDAQ:TSLA). Yeah, Tesla, the electric car company with no profits, a price/sales ratio of 13.75 and a price/book ratio of 41.49. For perspective, consider that the S&P 500 medians are 2.06 and 3.19 respectively and that the corresponding ratios for Amazon.com (NASDAQ:AMZN), long seen by many as the poster child of ridiculous overvaluation, are 2.09 and 15.99. As to P/E, we obviously can't compute it for TSLA since the company remains in the red. But still, TSLA stock remains far above the levels at which classic cash-flow-oriented valuation models might put it.

Figure 1

Obviously, when it comes to TSLA's stock price, there's a lot going on besides value. And because of that, according to what had long been one of my favorite investment guru-quotes from Peter Lynch, it would seem that the complacent longs are sitting ducks waiting to get their heads handed to them.

If I could avoid a single stock, it would be the hottest stock in the hottest industry, the one that gets the most favorable publicity, the one that every investor hears about in the car pool or on the commuter train - and succumbing to social pressure, often buys.

Hot stocks can go up fast - usually out of sight of any known landmarks of value, but since there's nothing but hope and thin air to support them, they fall just as quickly. If you aren't clever at selling hot stocks (and the fact that you've bought them is a clue you won't be), you'll soon see your profits turn into losses, because when the price falls, it's not going to fall slowly, nor is it likely to stop at the level where you jumped in.

- One Up on Wall Street (1989), page 149 of the 2000 Fireside Edition.

It sure seems like Lynch was talking about TSLA, doesn't it (the auto industry isn't hot, but electric autos sure are). He wasn't, since he penned those words long before TSLA came into being. But let's play "lit crit" (literary criticism) and say Lynch was foreshadowing TSLA.

So what, exactly, is going on with TSLA? The market isn't dumb. Lynch and his teachings have been out there for a long time. Ditto Warren Buffett and the Ben Graham revival he sparked. We know what value is. And for those of us who might have talked ourselves into believing it was pushed aside by some sort of new paradigm, one would surely have thought that the horrifying experience of 2000-02, not to mention the late-2000s real estate crash, would have taught us to disavow such wishful nonsense. And it's not as if anybody thinks TSLA is more prosperous than it is. This is the information age; we have all the numbers we want just a few mouse-clicks away and all available for the low, low price of zero. So we have to assume the investment community knows what the valuation metrics show.

Let's consider the core problem with valuing TSLA: Nobody knows how big the electric auto market will become, how long it will take to reach a normal level, and how big a share of it TSLA will have. The business is in the process of being invented. We have no relevant precedents and no reliable data we can use for valuation models. Lots of people have ideas on these topics and offer assumptions. But that's all they are, assumptions - or to put it less politely, guesses or even wild guesses. The truth, the whole truth, and nothing but the truth is that NOBODY KNOWS what TSLA is worth! Here, however, are some things we do know: TSLA logged $615 million in sales in the latest quarter, up 101% from the year earlier quarter, when sales were $306 million, and up at an average annual rate of 297% from the fourth quarter of 2011, when sales were just $39 million. None of those growth rates come anything close to anything a value investor can plug into a classic valuation model, which typical thinks in terms of growth rates that are normal, sustainable through infinity. We also can assume, with a reasonable degree of confidence, that electric autos really matter and that our ability to keep on driving without continuing to exhaust and ultimately non-renewable source of fuel really is a big deal. And there seems to be enough coming from people who've actually driven Tesla autos that they're pretty good, enough so to justify an assumption that the company will have at least some reasonable share of this market. Moreover, precedent from the business world, particularly manufacturing and technology, allows us to assume that as electric car production comes down the learning curve, margins and profits will eventually wind up at adequate levels. The problem, though, is that we can't really use any of these assumptions to come up with a reliable estimate of value for TSLA stock.

One of the best at valuing emerging companies is Professor Aswath Damodaran of the Stern School of Finance at New York University. On March 25, 2014, he published an article on Seeking Alpha explaining why he increased his TSLA valuation from $67 (according to a blog he published in September 2013) to $99.85-$118.47. That was still below the prevailing stock price, but the most noteworthy thing was the magnitude of increase in such a short time and his closing thoughts:

It is easy to become cynical about markets and to cast those who have different views than you do into the "irrational" or "crazy" camp. Even if the Tesla run-up turns out to be overdone, look at the bigger picture, which is that a company that was non-existent five years ago has shaken up a sector where there have been no new entrants for decades. I have nothing but admiration for what Elon Musk has built over these few years, and I hope he succeeds. In fact, I will keep valuing Tesla every few months, and there will be a time, sooner rather than later, where I know it will be part of my portfolio. Just not yet!

While this is obviously not a suggestion that TSLA should be bought, it's a far cry from the dire warning sounded by Peter Lynch, about stocks like this gong up out of sight fast and falling just as quickly.

Is This REALLY About Value?

I submit to you that none of this, not even Damodaran's work on TSLA, has anything to do with value.

We know what value is and we have several good, sensible ways to calculate it. One thing all such methods have in common is that they seek a rational, defensible, reliable relationship between the stock price and reasonably visible measures of corporate wealth creation (dividends, earnings, cash flow, etc.). Damodaran's effort, however sensible it may be, provides none of that. In his Seeking Alpha article, he offers an excellent Excel download detailing his computations but it is not valuation. It is guesswork. It rests heavily on ten-year assumptions about a business that offers no evidence from which we can even reliably make 10-month assumptions.

What Damodaran's spreadsheet is, in my opinion, is a story builder, a great story builder, the best story builder I've seen, a story builder anybody considering TSLA would do well to download and use, but a story builder nonetheless.

Here's another phrase for a story builder: a noise analyzer.

I want to use the latter phrase because it ties into the Robert Shiller-inspired framework I presented in my March 20, 2014 Seeking Alpha article entitled Overvaluation Is not Necessarily A Reason To Sell. In that article, I explained that we should not necessarily expect the price of a stock to equal its value but that in the normal course of market behavior, Price is equal to Value plus Noise. (In other word, P is not equal to V. Instead, P = V + N) I further explained that the role of noise depends on the level of "friction" an important element of which is information cost. When the information needed to reasonably value a stock is high or unavailable (i.e. the cost is infinite), then noise becomes more prominent and the importance of value diminishes, or possibly even vanishes.

In that article, I cited TSLA as a stock whose 3/17/14 price was based 100% on noise and zero on value. It's important to understand, here, that when I say value, I'm not referring to any spreadsheet exercise one can come up with. I'm talking about a highly credible assessment of value, one in which all market participants attempting to make such a calculation are more or less likely to agree (that being the essence of the efficient market theory, which assumes P is always equal to V). Many investors downloading the Damodaran model can make defensible and substantial modifications to its assumptions and come up with very different valuations, as Damodaran himself did between 9/13 and 3/14 and as he acknowledges he is likely to do as the months pass.

Bottom line: TSLA is all about noise, not value.

Noise Is Not Necessarily Bad

Up till now, I've been distinguishing between noise and value without suggesting whether high-noise stocks are worse or better than low-noise stocks. That will change now. (And stay with this; when all is said and done, you might be surprised at what it means for TSLA and other big-name high fliers.)

Given the current market culture, characterized by readily available information and the extent to which investors have had exposure to some measure of investment education (formally in schools and/or informally through books or seminars) and the widespread homage to the likes of Warren Buffett and Ben Graham, I'm going to start with an assumption that low-noise stocks are likely to outperform high-noise stocks.

To test this, I define low-noise stocks as those for which noise comprises less than 25% of market capitalization (the computations are introduced in my March 20th article, and delineated with considerable detail in its Appendix) and that high-noise stocks are those for which noise comprises more than 75% of market capitalization. I'm also going to look separately at the Russell 1000 constituents, a generally well-followed larger-cap group for which information costs (i.e. the burden of computing reliable valuations) ought to be low, and the smaller-skewing Russell 2000, which is likely to have more less-intensely-analyzed companies for which noise is likely to be prominent.

The Russell 2000 Group - On Script

A stock screening strategy that buys low-noise Russell 2000 stocks significantly outperforms a strategy that buys high-noise stocks from the same universe. Figure 2 shows the results of a backtest of the low-noise portfolio from 12/31/99 through 3/17/14 which assumed 4-week re-balancing and 0.25% price slippage on every trade. Figure 3 shows the backtest results for the high-noise strategy.

Figure 2: Low-noise Russell 2000 (smaller) stocks

Figure 3: High-noise Russell 2000 (smaller) stocks

I also tested ultra-versions of each strategy. Ultralow-noise stocks are those for which noise is less than 10% of market cap while ultrahigh-noise stocks are those for which noise comprises more than 90% of market cap.

Backtesting showed little difference between performance of the low-noise and ultralow-noise portfolios. In other words, once we cut noise down to 25% of market cap, we'd gotten as much out of this investment theory as we're likely to get. But it was different on the high end. Figure 4 shows backtested performance of the ultrahigh-noise portfolio.

Figure 4: Ultrahigh-noise Russell 2000 (smaller) stocks

On the whole, this is pretty much what I expected to see. When dealing with smaller stocks, it's a good idea to try to steer clear of market noise. Try to emphasize stocks that can be reasonably valued. This does not mean one should look only at stodgy low-P/E stocks. High P/E stocks are perfectly fine - if you can confidently justify the valuations through superior fundamentals, particularly those relevant to growth. Whether your valuation analysis points you to low or to high P/E ratios is immaterial. What is essential is that you have enough information to feel confident in whatever valuation you wind up using.

Now, let's turn to the world of the larger Russell 1000 stocks.

The Russell 1000 Group - What the $@%&!

I went into this thinking that as worthwhile as it could be to minimize noise among small caps, we should work even harder to do this for larger-cap stocks, about which we typically benefit from much more information and analysis. While I don't believe any segment of the market is truly efficient, I'm comfortable with the notion that the market for large caps is likely to be more efficient than the market for small caps.

A backtest of a low-noise Russell 1000 strategy gets this part of the study off on the right foot, so to speak.

Figure 5: Low-noise Russell 1000 (larger) stocks

Yep. That's what I thought I'd see; good market-beating performance from the low-noise group. The performance of ultralow-noise Russell 1000 stocks was about the same.

Now, let's look at the high-noise Russell1000 portfolio. Warning: This chart is for real; it's not an upload error.

Figure 6: High-noise Russell 1000 (larger) stocks

How about them big-caps! That's almost as good as what we saw for the low-noise large-caps. It's not quite as good, but it's close, and certainly a far cry from what Peter Lynch suggested we should expect. And interestingly, the performance of the ultrahigh-noise Russell 1000 group was even better.

Figure 7: Ultrahigh-noise Russell 1000 (larger) stocks

After scratching my head, contemplating my navel, sleeping on it, and double, triple and quadruple checking to make sure I didn't make some sort of mistake, I think I understand why my initial expectation (the one that was blown away by the test) was off base. I had assumed that the improved quantity and quality of information available on larger stocks would enhance the role of value and make it that much more dangerous to invest on the basis of noise. I think I was right in assuming valuation would be well worthwhile for larger-cap companies (see Figure 5). What I had missed, however, was a potential distinction between high-quality noise and low-quality noise.

Low-quality noise can be seen as investing on the basis of emotion, seat-of-the-pants nonsense or just plain idiocy (such as the 1960s, when investors fell in love with companies whose names ended in "onics," in the recent bubble, when any company whose name ended in ".com" was a star). Higher quality-noise entails genuine information and analysis, sometimes a lot of very sophisticated information and analysis, but information and analysis that deals with topics other than reliable, classic, familiar stock valuation.

The aforementioned Seeking Alpha analysis of TSLA published by Damodaran is an example of what I would call high-quality noise. As discussed above, although Damodaran presented his analysis using the vocabulary of value investors, I don't consider it genuine value within the meaning of this series of articles because it's highly subjective and involves much speculation about the future. Indeed, Damodaran expressly invites readers to plug in their own assumptions and share them via a Google-drive spreadsheet he made available (too bad nobody took him up on that as of the last time I looked; it seems like an intriguing community exercise). The sort of value I'm talking about here, value as differentiated from noise, needs to be much better defined than that. In the real world, disagreements will always occur, but with the value about which I speak, the range of plausible opinions should be reasonably narrow.

What I think Figures 6 and 7 are telling us is that high-quality noise, the kind that would seem more likely to be associated with larger caps (i.e., more intensive and sophisticated analytic coverage), can be every bit as bullish as value. In other words, you don't need to invest on the basis of value. But if you are going to instead rely on noise, or stories, at least make sure the stories are rational and sensible, as we often see right here on Seeking Alpha for such large-cap noisy stocks as Tesla, PLUG Power (NASDAQ:PLUG), Amazon.com, Facebook (NASDAQ:FB), etc. Not all stories published on such stocks are bullish. Some suggest the future won't be all that bright. Difference of opinion is normal. It's up to the investment community to weigh and balance the competing arguments and make its decision. What's not OK (as many shorts have discovered to their chagrin) is to be bearish on such stocks for no reason other than high valuation ratios or the absence of profits.

I should remind readers here that the tests necessarily deal with aggregates, not individual situations. I believe that in general noise is more likely to be of higher quality among bigger, better-followed stocks. But exceptions always exist: You may find individual instances of high-quality noise among small caps and low-quality noise among big caps. The challenge, if you're looking at an individual stock as opposed to following a quantitative strategy such as the one presented here, is to assess the quality of the noise. It's important to be objective. Resist the urge to dismiss a bullish story as being low quality simply because you started your inquiry with a bearish perspective. Conversely, try to refrain from buying into a bullish story simply because you already own the stock and feel a need to justify a decision you already made. The most dangerous thing you can do is evaluate the quality of a story based on whether it supports or refutes your own preconceptions. Aim for objectivity, and let Mr. Market lend a helping hand. If the market seems to be buying into a story, at least approach your assessment with an innocent-until-proven-guilty attitude. If the market is dismissing a story, then start with a guilty-until-proven-innocent perspective. You can disagree with Mr. Market, but do so after having objectively considered the story, not beforehand.

How Much Noise (Even High-Quality Noise) Is Too Much

It's important to avoid the temptation to assume that recognition and a willingness to analyze noise takes us completely off the hook with the burden of relating share prices to fundamental wealth creation. Since we are outside the realm of value, we shouldn't really try to say noise is OK but only so long as the P/E is below 50, 100, 200, or whatever. Even so, we still can't allow noise to become excessive.

The main way to judge whether we have too much noise is, simply, by the credibility of the stories being told. For example, it would be one thing to invest in TSLA based on assumptions about which you thought seriously such as those mentioned above and possibly input into the Damodaran spreadsheet (which would obviously have to differ from his, which presently suggest, at least to him, a no-go decision). But investing in TSLA simply because you think it's "the wave of the future" is an entirely different matter.

Another thing we can do is try to get a general sense of how prevalent noise is in the market as a whole. If we're willing to assume higher quality noise involves more diligence and effort, then we might be willing to consider the notion that as total noise mounts, we might be seeing increases in lower quality noise, the type that may prove more fickle, and hence, more dangerous.

Figure 7 shows the relative backtested performance of the Russell 1000 high- and low-noise portfolios, assuming a common index value of 1.00 for each as of the 12/31/99 start dates.

Figure 8

Many who bash TSLA and stocks like it take heart in the excesses we saw back around 2000, trying to liken these stocks to those that soared and crashed back then. Figure 8 suggests that the market was exceptionally enthusiastic about stocks for which noise comprised more than 75% of market cap. But it also shows that once that situation had quickly and harshly corrected, the excesses did not reappear.

We do have to take note of the upward slope we see in the relative strength of the large-cap noisy portfolio since 2009. This, of course, reflects the price action we've seen in TSLA, AMZN, PLUG, FB, etc. The slope is much too gentle and the degree of recent relative strength too small to cause us to panic over the prospect of a repeat of 2000-02. But the direction is sufficient to remind noise investors that they really need to be vigilant about the quality of the stories that motivate them, lest they start sliding down the slippery slope at the bottom of which we find pure emotion; low-quality noise.


Personally, I don't tend to play the big-cap noisy segment of the market, so I don't own TSLA or any other names like it. When I put my own money to work, I'm most likely to be in the small-, micro- or nano-cap portion of the market, using with bland ETFs for more general exposure. But that's me. Mr. Market, through TSLA's price action, is speaking loudly and clearly that there are plenty of others who do want to play this sort of game.

My suggestion for those who have a stake in TSLA or are considering one is to download Damodaran's spreadsheet (or find or create a comparable one), take it seriously, and be in or out of TSLA depending on your output. If you don't want to go through a process such as that, then I think you should respect the nature of the game and stay in the grandstand. Going into a high-noise stock like TSLA without some bona fide grunt work is like playing football without a helmet. You might turn out OK, but your chances of serious injury are much greater.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.