Description

This write-up isn’t about a stock but a segment of the market.

QUICK THESIS: Valuations for everything are terrible, but based on valuation, value stocks right now are the leper with the most fingers. We think growth stocks, on the other hand, have reached Jeremy Grantham’s definition of a bubble (100% above long-term mean). If you have a short book, we think you would do well over the next few years to pick short candidates within the most expensive 10% of the market. We’re not saying short something because it’s in the most expensive decile, but if you can find something shortable within it, you’re putting yourself on the right side of a Lollapalooza effect. That is the thrust of this write-up. I’m expecting this to get a low rating because it’s about broad market valuations and not a specific stock. Nonetheless I think this is important enough to post. 

If you’re like us, you think every time somebody makes a macro forecast, an angel gets its wings ripped out at the root. Macro is hard. Even Lord Keynes, on whose supposed wisdom our economy rests, couldn’t make any money at it. In the second half of his career, the successful part, he rejected macro and turned towards bottom-up value investing. You might be tempted to dismiss this write-up out of hand on the same theory. If I were you though, I’d at least take a look. Ask yourself if your strategy has correlated with what we show here.

Right now, everyone is talking about the valuation of the stock market. Frankly we don’t give a rip. The only thing that concerns us is the valuation of value stocks. Are they a good deal relative to historical valuations or aren’t they? And if they are expensive, we want to know how we protect ourselves.

To answer this question about the present, we first went back into the past. We used the Compustat database to look at the performance of value over two different market cycles: the 2001-2007 bull market for value and the 2007-2012 bear/sideways market. We did this to see what, if anything, those market cycles could tell us about where we are today. 

2001-2007 was one the best times in the last 40 years to be a value investor. This period also coincided with the hedge fund boom. Many of those hedge funds were value funds, and many of those value funds produced double-digit alphas over the market. Value indexes also absolutely crushed the market and growth indexes over this time (by 4% and 8% a year respectively). You didn’t even need an active manager picking stocks to benefit from buying cheap companies.

But just as 2001-2007 was the Woodstock of value investing, 2007-2012 was its disco era. According to Aswath Damoradan, most value funds didn’t outperform their indexes over that period. Meanwhile most growth funds did.  

It gets worse. From 2007 to 2011, growth indexes were up while value indexes were down. So for five years, most value funds not only failed to outperform value indexes, they also failed to beat the market and they failed to beat growth.

Annual return for the Russell 1000 (large caps) and Russell 2000 (small caps):

                                                                                                          Source: Russell.com.

As you can see, value underperformed for both the large caps in the Russell 1000 and the small caps in the Russell 2000. When value fails, it seems fund managers fail even more than the value index does itself. Intuitively that makes sense because funds charge a fee, and that drags down their performance relative to the fee-less index.

So what happened?

We find that what made value so effective in 2001 was what made it so ineffective in 2007. And what is that? Funny enough, it’s the very thing that makes value investing tick: valuations.

Using historical data from 2001 to 2014, we looked at the largest 50% of the stock market and screened out illiquid stocks and financial companies. The resulting stock “market” we were left with consisted of companies large enough for an institutional strategy. The smallest stock in this universe today would have a $275 million market cap, and the median would have a $2 billion market cap. The largest would of course be Apple. We then split this universe into deciles based on Price/Book in order to compare the cheapest decile with the most expensive. The point of these deciles isn't that you'd necessarily invest in them but that they give a nice clear view of how value stocks and growth stocks are valued at any given time.

Why did we use Price/Book? Because book value is relatively stable from year to year. The market’s P/E is a notoriously bad predictor of long-term stock market returns because the “E” is so volatile that it’s not a good number to key off. The Shiller P/E, which averages earnings over 10 years, is a great predictor of long-term market returns, however, because it stabilizes the “E.” We believe Price/Book, like Tobin’s Q, accomplishes the same thing.

Below is the median Price/Book of the cheapest 10% of the market (minus financials) from 1999 to 2014. The dotted line is the median Price/Book for those stocks (i.e. the value decile) over the entire period.

In 2001 the value decile was at one of its cheapest points in the entire 15-year period. If you started a value fund in 2001, you proceeded to look like a genius for the next six years. This was one of the best times to get into value because you were getting cheap companies not just at a discount to the market but at low prices in absolute terms.

In 2007, however, the value decile was at its most expensive point. The median Price/Book of the cheapest 10% of the market was 1.3, well above the 15-year median of .88. Essentially the run-up had already occurred, and since it couldn’t continue forever, it stopped. As valuations fell back to earth, bottoming in 2009 and again in 2012, value investors suffered—not terribly but enough.

What’s our takeaway here?

(1) Absolute valuations matter.

(2) Value investors’ returns correlate to systemic valuations far more than they think.

Having seen the power of absolute valuations, the next thing we did was take a look at relative valuations. We compared the median Price/Book of the cheapest 10% of stocks to the median Price/Book of the market. And here the plot thickened.

In 2000 and 2001, the cheapest stocks were cheap in absolute terms, but when you compare them to the rest of market, they were the deal of the century. They were trading at a 77% discount to the median market valuation. According to a separate study by Alliance Bernstein, global value stocks in 2001 traded at their biggest discount to the market in the last 40 years.

The tech bubble involved a terrific waste of wealth, and it was wonderful! If you were a believer in Ben Graham and thus were someone with a sense of history, this created a wonderful opportunity for you. Baron Rothschild famously said, “When there’s blood in the streets, buy property.” Well, thank goodness, it turns out there’s no need for blood. You just have to strike when the world has found a shiny new toy.

That said, as much as 2001 allowed you to buy cheaply into value investing itself, 2007 offered you the opposite. Only in 2005 were the cheapest stocks trading at a worse discount to the market. Look at the line from 2007 to 2012. That is the essence of mean reversion. Investors who bought “expensive” cheap stocks were treated accordingly.

Conclusions thus far:

(1) Absolute valuations matter.

(2) So do relative valuations.

Sanity check: EV/Sales

Any time you analyze historical information to look for patterns, there’s the risk you’re mistaking randomness for truth. As a sanity check, we performed the same analysis but with EV/Sales instead of Price/Book. Also instead of looking at the very cheapest companies, we looked at the cheapest 20%. Finally instead of looking at the largest 50% of the market, we looked at the largest 30%. And this time we included financials.

With EV/sales, you see almost the exact same thing you did with Price/Book. Value stocks were cheap in absolute terms in 2001—though they were even cheaper in 2009. Then their valuations peaked in 2007, which began the period of underperformance that only ended when valuations bottomed in 2012.

EV/Sales confirms the results of Price/Book.

What growth tells us about value

We now add a further layer into our analysis and see whether the valuation of growth stocks has any effect on the performance of value stocks. I know what you’re thinking. That reasoning is about as straightforward as Chevy Chase’s in Caddyshack: “In one physical model of the universe, the shortest distance between two points is a straight line…In the opposite direction, Danny.” Nah nah nah nah.

But remember: even though absolute valuations matters, investing is about alternatives. Nah nah nah nah.

In 2000 and to a lesser degree in 2001, growth stocks were at their most expensive absolute valuation ever. So when value stocks were at their cheapest, growth stocks were at their most expensive. This was great for value and terrible for growth. However, by 2007, growth stocks had fallen back to a more reasonable valuation of “only” 11x book value.

From 1999 to 2001, the most expensive 10% of stocks enjoyed their highest valuation relative to the market ever. However, by 2007, growth stocks were…I almost can’t bring myself to say it…they were cheap? Well, they were cheap compared to their 15-year median at least.

Recall that 2007 was the beginning of a terrible time for value. Now we know what happened. At the exact time that value stocks were at their most expensive relative to the market, growth stocks were at their cheapest relative to the market.

Also you may have noticed that absolute and relative valuations for the value decile were just as bad in 2005 as they were in 2007, and yet value continued to work for another two years. The difference between 2005 and 2007, however, was that in 2007, the discount to growth had become the smallest in 15 years. This shrinkage in the valuation gap strikes us as “the straw that broke the camel’s back.”

So in 2007 you saw three forces converging at the same time, in the same direction. Value suffered from:

(1) a high absolute valuation

(2) a small discount to the market, and

(3) a small discount to growth

When you see all three things, you’re observing one peaked-looking canary in the coalmine. Value investors should be prepared for low returns or even a sharp correction.

We think these results are fascinating. It seems value investing itself is subject to its own principles.

Where are we today?

Before we get into that, one quick word about how to use these valuations. They're useless in terms of timing the market in any given year, but they're useful in terms of looking at longer-term returns for an asset class. Our goal here is to put the trend on our side.

Warren Buffett has said interest rates act like "gravity" on stock prices. We think there' s a far more powerful force pulling down on stock prices, and that's valuations. But if you're looking for certainty about next year or something like that, look elsewhere. These valuations deal only in probabilities, not fixed outcomes. If you want perfection, start a doll collection.

Now onto the valuation of the market itself:

The market isn’t cheap. In fact in terms of Price/Book and EV/Sales, it’s higher than it was at the previous peak in 2007. It’s also not that far from the peak in 2000. The market at this point is objectively expensive.

Moreover the market over the last 15 years has hardly been cheap when you compare it to its long-term history. You may have been wondering if valuations over the last 15 years were simply trough valuations when viewed in broader context. (You and I both know you weren't actually thinking that, but it's something I need to address because otherwise everything I’m saying is nonsense). 

Check out this chart below of Warren Buffett’s favorite market valuation tool: Total market cap/GNP. As you see, even the median valuation over the last fifteen years is well above the previous peak in 1969.

Everything right now is expensive. And other than 2003 and 2009, it has been the entire time. Now as for value stocks, at least as of March 2014, valuations were on the expensive side in terms of absolute valuation. Fortunately though they’re were still attractively priced relative to the market.

In March 2014, the median Price/Book of the cheapest 10% was 1.10. The median over the last 15 years was .87. This suggests we’re entering a period of lower-than-average returns for value stocks.

EV/Sales confirms this:

Recall we included not only 20% of stocks instead of 10% here, but we also included financials and limited ourselves to larger companies. These results conflict with those of Price/Book, as they show value stocks as expensive. So which one is right? They both are. If larger value stocks and financials are expensive, then that suggests the best deals are more likely to be found in smaller companies and in non-financials. It’s not a guarantee of course or anything like that, but that is what seems to be the case.

Now here’s the good news (sort of). Even though absolute valuations for cheap stocks aren’t great and in fact are quite bad, the valuations for everything else are much, much worse. As the French philosopher La Rochefoucauld observed in the 17th century, "It is not enough to succeed. One's friends must fail." This comparative advantage might seem like small comfort. It’s not—particularly when you look at what is going on with growth stocks.

Growth stocks are at their most expensive in absolute terms since 2001. They’re also expensive even when valued against an already expensive market. For us, that puts the valuations of these companies in the danger zone. These stocks have reached the point where the lazy investor who stays in bed for the next 5 years will probably outperform the eager beaver aggressively trading them.

Below is the valuation spread between the cheapest 10% of the market and the most expensive. The bigger this spread is, the more it suggests growth stocks are overvalued and value stocks are undervalued--at least relatively.

The spread is the highest it’s been since 2001. In relative terms, that’s great for value and absolutely deadly for growth. For us, sticking with value stocks will be good if anything simply because you’ll do worse holding anything else. And as for growth stocks, the bubble has officially begun.

We think returns will be terrible for everyone over the long-run going forward, but they’ll be better for value investors, and they could be even better for those who can intelligently and opportunistically short growth.

Where today is relative to the last 100 years

Jeremy Grantham has said that you get a bubble when valuations are twice their long-term average, so I looked at the cheapest decile of the market and the most expensive decile as a gauge of how value and growth are valued today relative to their 100-year average.

Unfortunately I can’t get numbers for the value decile and the growth decile over the last 100 years directly. I don’t have that kind of data. I am, however, able to back into it using the data used previously, which was from 2001-2012. Over that time period, I figured out the average discount to the median market valuation for value stocks and the average premium to the market for growth stocks. Next I figured out the median valuation for the market in a given year going back 100 years and then applied the average discount or the average premium to get a guess as to the valuation of the value decile and growth decile over that time. Take it for what it is: a rough guess.

When I do that, here’s what I get:

Right now, the value decile is 65-70% above its 100-year average in terms of P/book and EV/sales. What makes me think these numbers are right? Because I got the exact same answer as the one Grantham gave in his last interview: namely that stocks are 65% overvalued. Note: Grantham was talking about the broader market, but I get that 65% overvaluation for the market itself too.

Yes, value stocks have the same degree of overvaluation that the market does, but value typically outperforms the market over time, which is why we believe it's the leper with the most fingers.

Another 30% run-up, however, and value stocks will have reached Grantham’s definition of a bubble. And when that happens, watch out.

But as bad as things look for value, the outlook for growth is much, much worse (fortunately for all of us). It’s 160% over its 100-year average in terms of Price/book and 133% over in terms of EV/Sales. The growth bubble, it seems, is more than official. It's terminal.

As long-in-the-tooth as value is right now, we aren’t quite at a point where we’re going into cash, but we’re close. Growth, on the other hand, can shoot up into another tech bubble. It can stay high for another few years. It doesn’t matter. Its valuation is so high that it's highly likely that it's about to begin its lost decade.

We think this is a shortable probability so long as you don't introduce path dependency into your shorting (i.e. you must figure out a way to cap the downside in the event of a run-up--by using puts, for instance). I mentioned shorting the Russell 3000 Growth index in the title. I had to put something there, but I don’t think that's the optimal approach: who wants to short an index when you can short crappy companies. That allows you to go long ideosyncratic risk as well as broader market risk--and now bubble risk.

Some shorts that involve hyper-growth companies and have been mentioned on this site are the following:

I’m hoping with this write-up to inspire you all to come up with even more.

Disclaimer: Nothing the author writes should be construed as investment advice or a recommendation to buy or sell specific securities. Please do your own research. While everything the author writes is factually correct to the best of the author’s knowledge, a lot of this is guesswork and is far more subjective and far more prone of error than it may seem. You are encouraged to notify the author of any mistakes or oversights in the comments section, but the author assumes no liability whatsoever for the accuracy of the information herein. Moreover the author undertakes no duty to update the information contained in this write-up. Do not rely on the information set forth in this write-up as the basis upon which to make an investment decision.

Catalyst