invest(igations)

What If Sector Classification Is All Wrong?

The Global Industry Classification Standards (GICS) classify industries into eleven different sectors, and do so on a purely intuitive rather than a data-driven basis. But once you look past that intuitive level, the sectors make no sense. What do Cigna and Amgen have in common, really? Yes, they’re both in the business of “health care.” But as businesses go, they couldn’t be more dissimilar. What does a semiconductor manufacturer have in common with an IT consulting service? They’re both in “technology,” but I can’t think of anything else. What does an airline, a defense contractor, a manufacturer of electrical equipment, and a human resources firm have in common? They’re all “industrials.” But they operate in completely different ways, and the market knows that very well.

So I started wondering if there were a better way to classify industries than by using GICS sectors. There are several ideas out there.

One is to use “themes.” Portfolio123 classifies all industries into five themes: macro economic (most industrials, transport, discretionary, most materials), population growth (staples, some health care, communications, utilities), special (energy, precious metals, airlines), financial (banks, insurance, real estate), and innovative (biotech, pharma, tech). This makes intuitive sense, but almost all of the GICS sector classifications (with the exceptions of materials, industrials, and health care) remain intact, so it’s more like a grouping of sectors than an attempt to rethink them altogether.

Some folks at Open Matters, a machine learning company, suggest that we classify companies according to whether they’re asset builders (make and sell physical things), service providers (use people to offer services), technology creators (generate and deliver intellectual property), or network orchestrators (facilitate transactions and interactions). I like this idea, but I haven’t seen anyone follow up with it and actually classify companies in this way. A similar idea, suggested to me by an RIA on a Portfolio123 forum, is to classify industries in five groups: input resources, business-to-business, business-to-consumer, infrastructure and government, and facilitators.

Another way to approach this problem is to look at industry factors such as gross margin, growth rate, regulatory barriers, and so on, and classify industries accordingly. This would be a massive project, and completely dependent on which factors were chosen.

So I’ve come to the conclusion that the least biased method of dividing industries into groups is to simply let the market do it for me. As one friend told me, “The market is the ultimate crowdsource.” One can gather the returns for each industry and then group them by how those returns are correlated.

So that’s what I did. I used the GICS to classify all stocks into one of sixty different industries. I then generated an equity curve for each industry from January 1999 to January 2019, based on equal weight per company (excluding stocks with extremely low liquidity) and rebalancing annually. I then created a correlation matrix (using Kendall’s tau as my measure of correlation). If two industries had market returns that were very similar, their correlation was close to 1; if their returns were extremely dissimilar, their correlation was below 0.

Once that was done, the problem was how to group the industries based on their correlations. After a lot of research and experimentation, I ended up relying on the k-means++ algorithm, which is a bit complicated to explain. However, let me try via an analogy.

Imagine a swarm of flies hovering various distances from a central object. That central object is the return of the market, and all the flies are the returns of the various industries. The returns of some industries—metals, for instance, or oil and gas companies—are going to be very far from the middle, because their correlation to the market is low. The returns of others—health care providers, insurance, machinery—are going to be quite close to the middle, because their correlation to the market is high.

What a good clustering algorithm will do with a swarm like this is to take the fly that is farthest from the middle and see what flies are closest to it, within reason. Those flies then get placed within a cluster. We then see what fly is next farthest from both the middle and that first cluster, and form a cluster around that fly. In the end, with a lot of iterations and difficult choices, you can create coherent clusters.

Now these algorithms tend to create clusters that are somewhat larger or smaller than I wanted. I wanted clusters that represented between 2.5% and 25% of the market as a whole when looked at historically. I had to continually adjust my parameters in order to get clusters of that size. I also had to take some of the outer clusters and remove them from my data in order to get any sort of differentiation for the inner clusters.

Here are the industry clusters I found, in rough order from least differentiated (closest to the market as a whole) to most (on the edge of the swarm). (I excluded REITs, MLPs, royalty trusts, and BDCs from consideration here, because those operate on such different principles than other stocks.) Please keep in mind that I clustered these industries based solely on their historical returns, not on any preconceived idea of how well they fit into distinct categories; in fact, I clustered them more or less blindfolded, without paying attention to what industries they were, just so that I could be objective. Also please keep in mind that industry clustering is never going to be very precise.

• PRIMARY. Staples, insurance, electricity, gas, water, road & rail, health care providers and technology, packaging. These are core stocks, commonly known as “defensive” stocks. They’re the products that you simply can’t do without, no matter how bad the economy is.
• SECONDARY. Chemicals, aerospace and defense, machinery, conglomerates, trading companies, professional services, health care equipment & tools, internet retail. These are complicated businesses that mostly depend on other businesses. Few of them are direct-to-consumer operations.
• BASIC. Paper, wood, clothes, stores, drugs, infrastructure, auto parts. This category is very similar to the “Primary” category above: these are also defensive stocks.
• COMPLEX. IT services, software, wireless, non-traditional utilities, construction. These are complicated goods that fall in between the “Secondary” category above and the “Edge” category below.
• SERVICE. Commercial and consumer services, finance, air freight, building and construction supplies, leisure. This is similar to the “Secondary” category except that it includes more direct-to-consumer businesses and, with an exception or two, is centered around the idea of providing services rather than goods.
• EDGE. Biotech, semiconductors, computers, electronics, media. These are cutting-edge businesses that depend upon innovation and creativity.
• CREDIT. Banks, real estate, distributors, cars, furniture. These are businesses that respond sharply to the credit cycle.
• SPEED. Internet, social media, telecom, communications equipment, airlines. These are businesses that depend upon and benefit from long-distance communication.
• EARTH. Energy, metals, electrical equipment, marine shipping. Stuff you get out of the earth, along with the means of getting it from one place to another.

(These clusters—their names and the industries included—are the expression of my idea about sectors and clusters, and are therefore protected by US copyright.)

Here’s how the clusters performed over the last twenty years:

Here are some possible uses for this new classification.

1. Cluster rotation. Keep track of the returns of each cluster and buy the ones with the strongest momentum or the most undervalued ones.
2. Cluster concentration. Invest in the clusters with the least volatile returns, or the clusters with the lowest beta, or the clusters with the highest alpha. (The least volatile clusters are, in order, secondary, basic, and primary, and the most volatile are earth, credit, and service. The clusters with the lowest beta are primary, credit, and secondary, and with the highest beta are edge, speed, and earth. The clusters with the highest alpha are primary, secondary, and complex, and the lowest are speed, earth, and credit.)
3. Recognize the differences between clusters in terms of investment strategy. The primary, basic, and credit clusters will respond best to a value strategy, while the secondary, edge, and service clusters will respond best to a growth strategy. Sentiment indicators are very important for the service and speed clusters and relatively unimportant for the credit cluster. You really want to favor tiny companies in the secondary cluster, while in the primary, basic, and credit clusters, size doesn’t matter as much.

So I propose that we replace the eleven GICS sectors with these nine clusters. I think they more accurately reflect the real differences between different industries.

My top ten holdings right now: ARC, PCMI, GSB, RLGT, CTEK, PERI, PFSW, NTWK, CLCT, PDEX.

CAGR since 1/1/16: 41%.

How to Avoid Value Traps

A value trap is a relatively cheap stock whose price simply doesn’t rise no matter how long you hold it. I’ve done a research study of factors that might characterize value traps, and here are the ones that I think provide the best warning signs. After going through these factors, I’ll tell you how I performed the study, describe some systems that incorporate these factors, give their performance, and recommend some value stocks that I think are not value traps.

1. Avoid companies who file late. The SEC has set deadlines for company filings; most companies must file a 10-Q within 40 days of the end of their fiscal quarter, though companies with a market cap of less than $75 million can take up to 45 days. After that, companies need to file a form NT (for “Non-Timely”), which will give them a five-day grace period. If a company doesn’t file a form NT or takes more than the five days given, it may face deregistration, delisting, and an inability to raise capital. In general, companies who are late with a quarterly or annual filing tend to do poorly over the next few months, lagging the S&P 500 by an average of more than 2% annually.
2. Avoid companies with negative or very low unlevered free cash flow. Unlevered free cash flow is cash flow from operating activities plus after-tax interest expenditures minus capital expenditures. If a company’s unlevered free cash flow is less than 2% of its enterprise value (the sum of its market cap, non-controlling interest, preferred equity, and debt minus its cash on hand), it’s going to be hard for it to grow enough to escape value-trap status.
3. Avoid any company whose negative cash flow from investing activities is a lot bigger than its cash flow from operating activities, or whose cash flow from operating activities is negative to begin with. A company may look like a good value, but if its cash flow from investing activities (which includes capital expenditures, acquisitions and divestitures, sales and purchases of investments and plants, and other investing activities) is negative $100 million while its cash flow from operations is only $50 million, it’s going to face a pretty tough year. To rank companies on this measure, I take the cash flow from operations, add it to the cash flow from investing (which is usually a negative number), and divide the total by the sales for the year.
4. Avoid companies with high balance-sheet accruals. Balance sheet accruals are the increase in the most recent year’s net operating assets (NOA) over the previous year’s, divided by the last two years’ average total assets. Anything over 5% is pretty dangerous, and represents a serious threat to future financing. If your NOA is growing, your proportion of cash and non-debt liabilities to total assets is shrinking, leaving you little breathing room to exceed investor expectations.
5. Avoid industries with poor unlevered free cash flow to enterprise value ratios. Right now the industries with the worst ratio of unlevered free cash flow to enterprise value are hospitality, construction, aerospace and defense, mining, food, construction materials, utilities, car manufacturers, air freight companies, communications equipment, trains and trucks, airlines, tobacco, and energy equipment. This list may change, of course, and certainly some of these industries have had better days in the past. But at the moment, they’re not the best industries to invest in.
6. Avoid companies whose price is extremely low compared to recent highs. Look at the company’s ten-month high one month ago, and compare that to its current price. If the latter is only a small fraction of the former (say, 60% or less), it’s going to take a good long time for the company to regain its former glory.
7. Avoid industries with strong negative momentum. Industries whose companies have, on the aggregate, fallen most sharply in price over the last year are probably best to avoid. As of early February, those industries are diversified telecom, mining, direct marketing retail, diversified financial services, consumer durables, distributors, car manufacturers, auto parts, pharma, biotech, energy equipment, and transportation infrastructure.
8. Avoid companies with low volume-weighted momentum. Divide the company’s VMA(15) by its VMA(200). If the result is below 0.8, that stock may well be in deep trouble.
9. Avoid companies with low projected EPS growth. In a previous post, I talked about five predictive signals for EPS growth: low profit margin or ROA, high EPS estimates, low accruals, strong momentum, and strong operating earnings growth. If those signals are mostly negative—i.e., if the company is showing high profit margin or ROA, low EPS estimates, high accruals, low momentum, and shrinking operating earnings—that’s a strong warning that the company is going to have a very hard time growing its earnings.
10. Avoid companies with a low ratio of gross profit to enterprise value. If using TTM numbers, I’d set my threshold around 0.15. Anything lower than that and the company is probably overvalued rather than undervalued.
11. Avoid companies that are issuing a lot more shares. Equity financing is far more expensive than debt financing, and a company that is issuing more shares rather than buying them back may be desperate for cash, leveraged to the hilt, or simply not fiscally foresighted.
12. Avoid companies whose trailing twelve months sales are either a lot higher or a lot lower than their sales the previous year. When looking at value companies, you want to see somewhat steady sales—ideally, sales growth should be close to the average for all companies. Companies with huge increases in sales are likely to disappoint investors in the near future; companies with huge decreases in sales are unlikely to easily make up for them. As a rule of thumb, avoid companies in the top or bottom decile in terms of sales growth (the most recent twelve months over the prior twelve months).

Here’s how I came up with these factors.

First, I created a universe of value stocks as follows. After establishing some minimum liquidity rules, I isolated those stocks that were in the top 50% of their industry in terms of book value to fully diluted market cap, sales to fully diluted market cap, and trailing-twelve-month net income to fully diluted market cap (in other words, the lowest 50% in terms of price to book value, price to sales, and P/E). They also had to be in the highest 50% of the entire universe in terms of EBITDA/EV, using a formula that adjusts extremely low or negative EV up to $5 million. That universe of value stocks included an average of 574 stocks during the 1999 to 2019 period, and today numbers 570 stocks.

I then ran a rolling backtest on Portfolio123, testing the six-month returns of these stocks if bought every week from 1999 to 2018. The average six-month return of one of these value stocks was pretty decent—7.22% compared to 3.49% for the S&P 500. I then tested about a hundred different factors by buying only the lowest quintile and measuring the six-month performance. The dozen factors above had the worst performance for that low quintile.

So now I’m going to limit this universe of 570 value stocks by excluding those that fail on one or more of the tests I’ve outlined above. I’m also going to exclude financial and real estate stocks because free cash flow and enterprise value are not great measures for those companies. That narrows our buying choices down to only 86 stocks.

The six-month performance of stocks that pass this rather complicated and strict set of screens is somewhat higher than the plain value stocks, averaging 7.93%, an improvement of 1.5% per year. If you include only companies with a market cap between $50 million and $1 billion and use the screen, your six-month return climbs to 8.42%.

I’ve always held that ranking is better than screening. A company might fail just one of these twelve screens but have excellent metrics on the others, making it a better buy than another company that has poor but passing metrics on all twelve. Screening is like simply requiring a student not to get F’s and judging her performance accordingly; ranking is like judging her on her grade point average. So if, instead of using the screen, you create a ranking system based on these twelve factors and just choose the top ten stocks, your six-month return climbs to 12.46%, for an annualized return of 26.47%.

But you can do even better than that. Instead of the strictly circumscribed value universe I’ve outlined, you can combine two ranking systems. The first will be the twelve-factor system I’ve just outlined, equally weighted; the second will equally weight the four value factors, except that I’ll be using forward earnings (the mean analyst estimates for the current fiscal year) rather than TTM earnings (which tend to be less reflective of a company’s potential). In total, that makes sixteen factors: the four conventional value factors weighted at 12.5% each, and the twelve value-trap-avoidance factors weighted at 4.17% each. My top-ten ranked stocks now earn 16.58% every six months, for an annual return of 35.91%. (Of course, one should take transaction costs into account; if one assumes that each transaction costs 75 basis points, the return falls to 15.08% every six months, for an average annual return of 32.43%. Also keep in mind that average annual returns are always significantly higher than compounded annual returns.) For comparison’s sake, if you had created a simpler ranking system with none of the value-trap factors and only the four value factors and chosen the top ten stocks, your six-month return would have been only 7.66% because you’d be buying a huge number of value traps.

Now it’s not very practical to check this every week and hold the top ten stocks for exactly six months. For one thing, when prices are very volatile, you’ll have a very large number of stocks and no cash to deploy, while when prices are not volatile at all, you’ll be holding the same stocks for a long time with lots of cash on hand. For another, a six-month sell point is completely arbitrary. Why put in so much effort into choosing which stocks to buy and use an arbitrary measure to choose which stocks to sell?

So here’s a system. Each week, buy the top ten stocks by rank, assuming you don’t already own them, apportioning 5% of your portfolio to each one. Only sell them after you’ve held them for three months and if they’re no longer in the top twenty-five.

Assuming transaction costs of 75 basis points, your compound annual growth rate over the last twenty years would be 32.07% and your average annual return would be 37.73%, while you would be leaving an average of 12% of your portfolio in cash. You would beat the S&P 500 in fourteen out of those twenty years. You would hold your stocks for an average of four months, and your average annual turnover would be 2.5X. Your maximum drawdown would be 50% (compared to 55% for the S&P 500), and your alpha would be 27.25%.

So what are these ten value-but-no-value-trap stocks right now? Resolute Forest Products (RFP), Fujitsu (FJTSY), L.S. Starrett (SCX), Tech Data (TECD), Perion Network (PERI), GameStop (GME), NetSol Technologies (NTWK), RWE (RWEOY), Corus Entertainment (CJREF), and Veon (VEON). I somewhat foolhardily predict that only one or two of these stocks will go down more than 20% in the next six months, and that their total returns will beat those of the market as a whole. We’ll see if I’m right in six months’ time . . .

My ten largest holdings right now: ARC, GSB, PCMI, CTEK, PERI, PFSW, OSIR, PDEX, HALL, CLCT. I also own a fair amount of SCX and NTWK.

My CAGR since 1/1/16: 40%.