Value Is Fine But Low PE Is Struggling (I Mean It; No Typos)

I mean exactly what the headline says. Value is fine. It always was. It is. It always will be. The reason why you might think otherwise is because Value is often assumed to mean low P/E, low Price/Sales, low Price/Book, etc. That’s not necessarily so. I present here a screen that finds value stocks that work in today’s Goldilocks market.

on sale

What We Think We’re Seeing With Value

Superficially, it looks to many as if Value is cold. We all have our stories and our evidence. Here, for example, is just one of countless things that can be shown to support a value-is-cold argument. It’s a simple Portfolio123 five-year backtest showing a screen designed to find only one ETF, the iShares Russell 1000 Value ETF (IWD), and comparing it to a benchmark consisting of the regular Russell 1000 ETF (IWB). (The Value ETF is the red line.)

Figure 1

Fig 1 - value v R1000

That speaks for itself, right? No, not necessarily,

A Broader View of What We’re Seeing With Value

Table 1 goes further and compares the performance of the Value ETF, over the past five years, not just to the overall Russell 1000, but also to a couple of other ETFs; the iShares Russell 1000 Growth ETF (IWD), and the iShares Edge MSCI USA Momentum Factor ETF (MTUM)

Table 1

5 Year Test Value ETF Russ 1000 ETF Growth ETF Momentum
IWD IWB IWG MTUM
Annual % Return 9.63% 13.12% 16.38% 17.66%
Standard Dev. 10.61% 10.49% 11.19% 11.25%
Max Drawdown -16.10% -14.67% -14.79% -13.82%
IWD Tallies Relative to Benchmarks
Correlation – – 0.97 0.87 0.76
Beta – – 0.98 0.83 0.72
Annual Alpha % – – -2.56% -3.00% -2.36%

See what’s happening? It’s not as if the Value ETF is really weak. A 9.63% annual rate of return achieved over a period characterized by very low interest rates and very low inflation is nothing to be ashamed of. What’s really happening becomes apparent when we look at which other ETFs have been hotter, much much hotter, the Growth ETF and the Momentum ETF.

That market isn’t so much telling us it hates Value as it is saying it’s feeling feisty; it’s in love with bolder riskier stocks. For that message to make sense, we have to presume the market loves what it’s seeing from the economy and from corporate profits. And the data from only the last 12 months suggests Mr. Market is even more optimistic about future economic developments and profit prospects. (Apparently, investors are less worried about trade wars than are many who talk to reporters.)

Table 2

Last 12 Months Value ETF Russ 1000 ETF Growth ETF Momentum
IWD IWB IWG MTUM
Annual % Return 8.10% 16.62% 25.08% 26.00%
Standard Dev. 9.12% 9.45% 10.60% 12.07%
Max Drawdown -10.12% -9.57% -9.16% -10.92%
IWD Tallies Relative to Benchmarks
Correlation – – 0.94 0.80 0.76
Beta – – 0.91 0.69 0.57
Annual Alpha % – – -5.64% -7.46% -6.40%

By the way, Table 3, comparing a screen that picks up only the standard equity-income ETF, the iShares Select Dividend ETF (DVY) with a yield of 3.15% and compares it with a benchmark consisting of the Vanguard Dividend appreciation ETF (VIF), an ETF that accepts a much lower yield (1.83%) in anticipation of stronger dividend growth in the future, presents a different perspective on the same theme.

Table 3

Standard Eq. Inc. Div. Growth
DVY VIG
5 Year Test
Annual % Return 11.46% 11.12%
Standard Dev. 9.11% 9.85%
Max Drawdown -13.21% -12.64%
Correlation 0.86 – –
Beta 0.79 – –
Annual Alpha % 2.99% – –
Last 12 Months
Annual % Return 9.20% 15.34%
Standard Dev. 7.58% 9.23%
Max Drawdown -8.53% -9.34%
Correlation 0.88 – –
Beta 0.72 – –
Annual Alpha % 0.90% – –

While the five-year test shows the two ETFs to have been neck and neck, the one-year test reinforces the notion that Mr. Market has, more recently, been focusing intently on growth, that being the notion that could justify the much-lower-yielding VIG outperforming as it has.

Goldilocks

So now we know. Notwithstanding the sudden and surprising spike in the CBOE Volatility Index (VIX) that frazzled nerves and pulverized leveraged short sellers a while back, the market remains very much in Goldilocks mode, a scenario in which everything is just right (business trends are sufficient to keep profits rising at a nice clip without being so good as to cause troublesome increases in inflation or interest rates).

goldilocks

And in this environment, we should not be at all surprised to see disdain for companies perceived by Mr. Market to be dogs, or worse. 

That is exactly what we’re seeing with stocks that have low tallies for P/E and other value ratios. In this environment, they are supposed to underperform. And they have been underperforming.

So Why Do I Say Value Still Works?

The answer is in the complete definition of value. Low P/E is not it. Neither is low EBITDA/ED, low PB,  low P/S, low P/FCF, etc. Nor is it high BM (the academicians way of changing low price-to-book to high book-to-market without recognizing that the acronym really stinks).

For the sake of clarity, let’s just focus on low P/E. (The same ideas apply to all the other metrics.)

Low P/E is not, never was and never will be inherently good. Low P/E is only good if its lower than it should be given the company’s quality and prospects for growth. 

If you want to skip the details and jump to the grand finale, here it is: Rising growth expectations and/or diminished perceptions of business risk push fair and ideal P/E upward, and vice versa.

For those who want to go step by step, here’s the logic:

  • A stock should be priced at the present value of the investor’s expected future cash receipts.
  • For convenience and to adapt to uncertain resale assumptions, this has been restated as the Gordon Dividend Discount Model, which tells us P = D/(R-G) where P is price, D is the next dividend, R is the required rate of return and G is the expected dividend growth rate.
  • To trim some ivory from that tower and get us closer to the real world, I restate it as P=E/(R-G) where E stands for Earnings.
  • Harkening back to Algebra 101, we know we can divide both sides of that equation by E.
  • This results in the formula for P/E: P/E = 1/(R-G)
  • Taking our old math a bit further, we see that because G is a negative number in the denominator of a fraction we know that as G rises, so, too, should P/E (just like proponents of the PEG ratio have been saying).
  • We also see that R is a positive number in the denominator so that as it rises, P/E should fall.
  • R rises as interest rates and/or as market-wide perceptions of risk rise; but with respect to company-specific factors, R falls (causing P/E to rise) when perceived business risk declines.

For more on this, check my stock strategy design cheat sheet. 

Value still works because if we look for low P/Es (and low tallies on other value ratios) among companies whose growth prospects and/or risk characteristics suggest the shares deserve higher ratios than they have, we can succeed. 

It’s not about low P/E. It’s about undeservedly low P/E.

Empirical quant data showing that, over time, low P/E has worked is not so much about virtues of low P/E but about the phenomenon known as mean reversion. If we naively focus on a group of low P/E stocks, theory tells us we’re looking at a lot of middling to weak companies. Mean reversion refers to the cyclical nature of just about everything and tells us that in a world where flux is the norm, weaker companies are more likely than not to get better over time. Mean reversion also leads us to assume a naively selected group of high P/E stocks is likely to include many powerful growth firms destined to slow down as time passes.

For now, however, the market has thrown the mean reversion concept out the window. The market is assuming that what we’ve seen in the past represents what we’ll see in the future. That, essentially, is what momentum is all about. So now, with mean reversion out of sight and out of mind, if Mr. Market is seeing a low P/E, the stock is deemed guilty until proven innocent, and proof of  innocence requires evidence showing better than generally appreciated growth prospects and/or a better than generally appreciated company risk profile.

Screening for Goldilocks Value

Starting with the obvious, I’m going to sort stocks based on the Portfolio123 Basic: Value ranking system, a plain-vanilla model that looks a lot like many other ranking systems out there. (Value investing is not about super-top-secret formulas nobody else thought of; we all know and can easily calculate the same ratios. It’s about how the ratios are used. If you want to see details of this particular ranking system, click here.) 

I’ll select the to 15 stocks (the 15 that have the lowest ratios) and refresh the list every three months.

But I won’t stop here. If I do, I’ll wind up with a collection of stocks that’s likely to be no better than the ones causing everybody else to complain that value isn’t working any more. To go beyond that, I need to first prequalify my list to eliminate, as best I can (the future is always unknowable so we do our best approximating using the information we have today), stocks that don’t deserve to be valued cheaply as they are because the companies have better growth prospects and/or are of better quality (i.e., lower risk) than would justify such low ratios. In other words, before I rank for value (low P/E etc.) I screen for stocks that don’t deserve to be cheap.

Here’s how I do it.

  • I start with a Portfolio123 universe that approximates the Russell 1000 constituent list, a large-cap group. This alone provides some measure of company quality (diminished business risk) based on scale (the superior ability of companies to absorb fixed costs and to benefit from internal diversification resulting from different kinds of business operations).
  • It’s hard to screen for future growth. All we can do in our non-psychic world is work with indirect information that can be used as proxies for investor judgments in this case, favorable judgments regarding future growth prospects. For this screen, I require ranks above 80 (on a 0 to 100 scale) in terms of an academic-type momentum ranking system I sometimes use that eschews short-term fidgeting and looks to a longer time period.
  • I then add two-more safety valves designed to catch potential dumpster fires that may have slipped through the cracks. One such test eliminates companies whose score on the Portfolio123 Basic: Sentiment is below 25 (out of 100). The other safety valve test is a bit of an oddball. I eliminate stocks that rank above (yes, above!) 98. So for the cheapest of the cheap stocks, the potentially smelliest of the cigar butts, I’m not even going to consider anything else. I’m just going to assume we’re dealing with garbage and dump them down the chute before I sort to find the cheapest 15.

Testing the Screen

Table 4 shows the results of a basic start-to end backtest of the screen (it assumes a 0.25% slippage penalty on all positions that enter or exit the hypothetical portfolio). Table 5 shows the results of a rolling backtest test, in which a set of self-contained three-month portfolios each starting a week apart are tracked).

Table 4

Portfolio IWB
5-Year Test
Annual % Return 16.72% 13.12%
Standard Dev. 13.91% 10.49%
Max Drawdown -17.30% -14.67%
Correlation 0.75 – –
Beta 0.99 – –
Annual Alpha % 4.22% – –
Last 12 Months
Annual % Return 15.69% 16.62%
Standard Dev. 11.88% 9.45%
Max Drawdown -14.76% -9.57%
Correlation 0.72 – –
Beta 0.91 – –
Annual Alpha % 5.94% – –

Table 5

Avg. of 13-Week Tests Average Return %
Portfolio IWB Port. Excess
5-Year Test
All 3.67% 3.10% 0.57%
Market Up  5.40% 4.60% 0.79%
Market Down -4.16% -3.70% -0.46%

The results are in line with expectations. The numbers are somewhat OK (the five-year test works but the one-year test shows that the market has become even more buoyant over the past 12 months). Table 5 confirms the Goldilocks character of this Value screen by showing it’s not something we would want to use if we concerned about the future.

The Companies 

Here are the companies, in order of Value rank (lowest ratios on top):

Macy’s (M)

Micron Technology (MU)

Poplar Inc. (BPOP)

Holly Frontier (HFC)

US Foods Holding (USFD)

Acadia Healthcare (ACHC)

Centene (CNC)

CACI International (CACI)

Liberty Media Sirius (LSXMK)

Advance Auto Parts (AAP)

Brunswick (BC)

First Data Corp. (FDC)

Urban Outfitters (URBN)

HD Supply (HDS)

Vistra Energy (VST)

Conclusion

The screen is good for what it is, an approach to picking value stocks in a highly bullish environment in which the market thinks the future is rosy and has little use for companies that don’t feed into this world view. Because I am somewhat cautious, I would not use this as a pure model-based portfolio through which I’d own all stocks that appear. I’d use it instead as an idea generator that prompts further company-specific examination in which I’d want to know if the stock can satisfy a preference for attention to value. Even in today’s climate that would leave me less exposed than a pure momentum play would if clouds appear on the horizon. 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s