Horizon Kinetics commentary for the first quarter ended March 31, 2016.
H/T Value Investing World, Chris Mayer
- Part I: Reminder of how dangerous the ‘ETF Stock Market’ is and of the enormous risk/value tradeoff between index securities and individually selected stocks.
- Part II: Why Beta, the go-to fund selection statistic, is entirely misleading – and might lead you off a cliff.
- Part III: The cash balances and what we’re doing with them. Also, Core Value’s return to ‘equity bonds’ – not the 2002/2003, but the 2016 vintage.
Horizon Kinetics – Part I: Let’s Be Clear About Valuation Risk
Horizon Kinetics’ 4th Quarter commentary established just how much at risk the broad stock market is. Dominated by companies favored by the major indexes and ETFs for their trading liquidity and their low price volatility statistics, they serve the needs of ‘industrial scale’ investing. For the mature companies representing the largest portions of the S&P 500, margins are at historic highs, growth has slowed or stopped, and the ETF demand for the shares has inflated valuations to excessive levels.
Warren Buffett’s Annual Letter: Mistakes, Buybacks and Apple
Warren Buffett published his annual letter to shareholders over the weekend. The annual update, which has become one of the largest events in the calendar for value investors, provided Buffett's views on one of the most turbulent and extraordinary years for the financial markets in recent memory. Q4 2020 hedge fund letters, conferences and more Read More
The message should not be forgotten just because a few months have passed. So, we’ll supplement the discussion with a newly popular type of ETF: the low-volatility ETF. The iShares MSCI USA Minimum Volatility ETF (ticker USMV), for instance, has almost $12 billion of assets3, more than twice as much as last July. By comparison, the old standard, the iShares S&P 500 ETF (ticker IVV), has only 3% more assets than a year ago. The website ETF.com describes USMV thusly: “The fund typically overweights defensive, dividend-paying sectors … The ETF takes substantially less risk than the market portfolio as shown by low beta.” Really? Language has power, power to distort thinking and decision making. Part I of this review will compare “risk” and opportunity through the prism of formulaic investing, as via USMV, versus individual security analysis and selection, as via Core Value.
The Consumer Brand Group There are 168 companies in USMV, with several major sectors, which are just ‘safer’ subsets of the S&P 500. A large portion is the well-known brand companies, such as McDonald’s, General Mills, Coca-Cola, and Procter & Gamble. Now demonstrably mature, their revenues are either declining or essentially flat. McDonald’s, the 3rd largest holding, is an example, and the accompanying table provides some of McDonald’s financial data for each of 2008 and 2015. Here are two stories about that data.
One Story: Revenue in those 7 years rose by less than 1% per year, 7.98% in all, net income by even less. The balance sheet was more dynamic: property and equipment rose 14%, but long-term debt rose by over 135%. Its shareholders’ equity contracted by close to 50%. If you could have been told this story in 2008, would you have purchased the shares? Should we have?
Another Story: The McDonald’s share price appreciated by 90% over the 7 years, which is 9.6% annually.
Explanation: One thing that happened is that McDonald’s P/E ratio expanded from 16.9x to 24.6x. That shareholders were willing to pay more for the same earnings accounted for about half of the stock return. Another thing that happened was that interest rates dropped; for 10-year Treasuries, from 4.08% at the beginning of 2008 to 2.30% at year-end 2015 (and to 1.77% now). This permitted McDonald’s to finance a massive share repurchase program that would have been unaffordable but for these artificially low rates, which is why its interest expense only rose by 22% even as its debt ballooned by over 6x this amount.
Whether or not interest rates increase, though, it must surely be realized that the share price appreciation of the past seven years is hardly likely to be repeated.
And Another Story: An alternative might have been Wendy’s. According to etfdb.com, of the well over 1,000 U.S.-traded equity ETFs, none has significant exposure to Wendy’s; 29 have significant exposure to McDonald’s. Wendy’s was purchased in Core Value in 20124 not because it was the most profitable restaurant chain (that would be McDonald’s), but because it was about the least profitable.
In 2012, Wendy’s net profit margin was less than 3%. Domino’s margin was 6.7%, Panera’s 8.1%, and it was 11.8% for Yum Brands (the KFC, Pizza Hut and Taco Bell combo), let alone McDonald’s 19.8% margin. Wendy’s, though, had a plan to induce franchisees to remodel their stores. Experi-ments at company-owned stores had already demonstrated higher customer traffic, revenues and profit margins. A related strategy was to sell most of the company-owned stores to franchisees. Restaurant chains’ owned stores have a lot of revenue but very low profit margins; revenues from franchisees are akin to royalties: much lower but with extremely high margins.
By 2015, Wendy’s net profit margin more than doubled, to 8%. Between 2012 and today, the share price rose 150%. And the remodeling program is only one-third done. As it approaches half-way, remodeling expenditures are expected to diminish, while the earnings contribution from the growing proportion of reimaged stores will increase. Accordingly, Wendy’s expects free cash flow in 2018 – after capital expenditures – to be 33% to 66% higher than 2015’s net income. That’s about a 10% to 18% growth rate.
The Telecom Group USMV also includes top 10 positions in AT&T and Verizon (numbers 2 and 6), but one can’t watch television without seeing competitors advertising for customers to switch carriers based on deeply discounted price plans. Cellphone service is a classically high-fixed-cost, low-marginal-cost business: if the incumbent service providers are forced to lower their prices, there will be a virtually undiluted impact on operating income.
REITs and Electric Utilities Two major USMV sectors are REITs and electric utilities. Marketed as “bond substitutes,” they exhibit low price volatility care of declining interest rates and continuous buying by index funds. They are at all-time high valuations. The utilities trade at a remarkably low 3.5% dividend yield. These different sectors do not represent diversification. They simply represent different, very long-dated exposures to interest rates at an historic low—or put differently, additional ways to suffer severe losses if long-term rates rise. Moreover, elec-tric utilities face a rapidly expanding threat from the popularization of rooftop solar panels. Just a couple of percentage points of demand deficit represents tipping point issues relative to utilities’ ability to expand earnings at all or to earn sufficient returns on high-cost historical plant and equipment to support dividends.
As to the REITs in USMV, they yield only about 3%, or more than 30x earnings. The largest U.S. REIT is mall operator Simon Property Group (ticker SPG). There are 22 ETFs in which SPG is a significant holding. Real estate companies are not exempt from competition or economic cycle risk. SPG experienced a 1% decline in occupancy in 2015, while tenants’ sales per square foot rose only 0.1%. That is the wrong direction for a high-valuation stock with growth driven by rising base rents, overage rents and leasing spreads. The largest holding in USMV is Public Storage REIT (PSA). The self-storage companies have been expanding so rapidly via new construction and acquisition that the issue of market saturation arises. Their earnings growth has been fueled by 6% rent increases and large quantities of debt. This pattern, for which exaggerated valuations are paid (28x cash flow (FFO) and a 2.5% dividend yield for PSA), is not sustainable; it will probably end badly. Understand that Horizon Kinetics – right or wrong, one can’t know, yet – recently issued a short-sale recommendation on Public Storage REIT for our institutional research subscribers. The qualitative point (more about this in the Beta discussion) is that the superficial descriptive characteristics of large market cap, high trading liquidity and low recent volatility are not representative of low business or valuation risk. They’re merely a description of a package designed for ETF consumption.
Another Way: It is possible to get exposure to publicly traded real estate that is actually well below fair value and not nearing growth limits. We previously reviewed the deep discounts at which companies like Howard Hughes and Texas Pacific Land Trust5 trade, and the return possibilities on their underdeveloped properties. Last month Howard Hughes sold some properties it had assembled as part of its South Street Seaport development in lower Manhattan. It sold these buildings for $390 million, a gain of $140 million. Bear in mind, the South Street Seaport redevelopment is ongoing and remains far larger than when Howard Hughes came public in 2010, when the Seaport was on the books for $3.1 million. With development expenditures, the book value is now about $225 million, and a conservative estimate of its developed value is roughly $1 billion, net of development costs. (This valuation and book value encompass only 362,000 square feet of development, with 700,000 square feet of additional space left to be developed.) When we originally acquired shares of Howard Hughes for Core Value, we assigned zero value to the Seaport.
Horizon Kinetics – Part II: Examining Low Betas – They Absolutely Will NOT Protect You From Volatility
We just characterized the iShares Minimum Volatility ETF (USMV) with qualitative and contextual analysis: excessive valuations vulnerable to collapse in the face of slowing or stagnant earnings, rising interest rates and competitive incursions. But the market just uses a statistical assessment, so let’s try that: an equity beta of 0.69 and a standard deviation of 9.01% (as of March 31st), as against a 3-year annualized return of 15.4% (to December 31st). That return is only a touch higher return than the S&P 500, but, crucially, with lower volatility (S&P’s beta is always 1.00 and its standard deviation was 10.95%). Given that USMV is comprised of brand name companies, and given the statistical evidence of low price volatility, it is almost universally acknowledged to be ‘safe’ and ‘low-risk’.
So, what is beta’s value to an investor? It would have to be some predictive information about the risk of a future price decline, right? But all that beta can actually assert is that each of these equities declined by less than other equities during weak markets during the measurement period.
So, if the beta is based upon a 3- or 5-year history, does that capture the volatility behavior of REITs between 2007 and 2009, when even the best real estate companies, like Simon Property Group, fell 70%? Is that in the beta? The Simon Property Group beta (5-year) today is 0.59. The beta at the end of March 2007 was an incredibly low 0.3 – only 30% of the volatility of the S&P 500. From that day forward, the two years April 2007 through February 2009, Simon Property Group shares fell almost 70%, while the S&P fell about 45%. So, what does beta mean to you now?
Utilities are almost 9% of USMV. How can historical price statistics capture the future impact of price competition in cellular telephones or from solar power if it never occurred before, though the early stages are quite visible? Many homes and businesses are already leaving the transmission grid, enabled by more efficient solar cells and improved battery storage technology. Another obvious risk to electric utilities is interest rates. As a pure mathematical reality, because the absolute levels of rates and dividend yields are much lower than during the last substantive interest rate increase period, utility shares would experience far more price variability when the next rate increase occurs. That is not captured in the historical price data either. Moreover, electric power generation was historically a rate-of-return regulated business, which also protected them. For many utilities, that’s no longer the case. So there seem to be a lot of observable risks that are clearly not captured by the primary risk measure, beta.
Now a few fun examples of beta risk. As ETF fees for conventional indexes like the S&P 500 have been diminishing toward zero (the iShares Core S&P 500 ETF is down to 7 basis points, 7/100ths of 1%, which the providers cannot live on), the ETF organizers have devised more differentiated ETFs that can charge higher fees. However, it is almost impossible to launch a new fund if it does not back-test well. That means that it must have a low beta. So, name a major index, and odds are it will now have a low volatility version.
Here are the top 10 low-volatility ETFs. Interestingly, in the iShares Emerging Markets Minimum Volatility ETF (EEMV), Taiwan has a 17.8% weight, and South Korea has an 11.4% weight. Consider that from a risk point of view. EEMV, by the way, has a beta of 0.85. Is it just me, or is it weird that an emerging markets index has a lower beta than the S&P 500? How do they do that? (We’ll provide one answer in just a minute.)
- If there were some incident in which China decides not to acknowledge or respect the autonomy of Taiwan, would this 17.8% weight in Taiwan remain non-volatile? Probably not.
- It is well known that China is building artificial islands in the South China Sea from which to project military power. What if it attempts to assert sovereignty, via military threat, over certain sectors of the South China Sea over which other nations claim sovereignty? Over 72% of the ETFs assets are in East Asia. Is this a low-probability scenario? Would the share prices of this sector remain non-volatile?
- If North Korea were involved in some sort of military altercation with South Korea, would the 11.4% position in South Korea remain non-volatile? Would the other East Asian markets react? Is this a low probability event? Consider that the dictator of North Korea is generally assumed to be quite unpredictable.
- Finally (although we could go further), there is the right-hand column of numbers in the preceding table of the largest low-volatility ETFs. Those are the weightings of each fund in financials. Each and every one has a large position, including USMV and the just-reviewed EEMV. Both the PowerShares S&P MidCap and SmallCap Low Volatility ETFs have in excess of 50% weightings in financial companies.
A rhetorical question: would an active manager of a low-risk strategy be permitted to take the risk of a 50% weighting in financials? Nevertheless, these portfolios can legitimately be characterized as low volatility, since of late the financial sector has not been volatile. And the high weighting assists the ETF in attaining its advertised low beta. But is low volatility an inherent attribute of companies in the financial sector? Or perhaps it is simply the case that the central banks of the world have maintained a stable, low rate environment for a very long time. Would anyone legitimately assert that these ETFs will remain non-volatile if rates rise?
One can see how the use of historical statistics actually distorts the risk analysis problem. As well, it distorts the ETF creation process.
Horizon Kinetics – Part III: Core Value’s Cash – Fortune Favors the Well Prepared
I hope we’ve provided some evidence that standard deviation and beta are merely ex post facto calculations. Rather than ask what the beta of USMV is, perhaps one should ask what it will be. If the goal is to avoid volatility, one must identify risks before they materialize. Investors who buy low-volatility funds are, in our opinion, unfortunately, really buying the next high volatility fund. It is also an unfortunate fact that the traditional Graham & Dodd value metrics, such as low P/E, low price-to-book ratio, companies with net/net balance sheets, or liquidation value investments, do not back-test well. This is because in order to have these desirable attributes a security must first exhibit poor performance and high volatility. Since one cannot market a back-tested index of poor performance, the traditional metrics and investment opportunities are discarded. One can’t market a deep value or opportunistic ETF. Really – go ahead and try.
Which brings us to the high cash position in Core Value. It was created as we reduced positions that had become substantially overweighted by virtue of appreciation and for which the return/valuation tradeoff was no longer distinctly in the client’s favor. We also reduced or eliminated positions that had become obviously overvalued, or for which the margin of safety had deteriorated. This was done gradually, over many months. Examples would be Jarden, which had a become 10%+ position but also well above its original single-digit P/E multiple, or Liberty Media QVC, which was trading at 22x earnings and, at the end of the day, is merely a retailer. So these were not market timing decisions. Nor, as discussed previously, was the cash intended to protect a portfolio during the market downturn that we anticipate, because it won’t really make that much difference. The cash is to provide the flexibility to capitalize on the opportunities that present themselves every handful of years or every cycle. True opportunities. The kind that give multi-year benefits.
See full PDF below.