Horizon Kinetics commentary for the third quarter 2015, titled, “The Long Road To The Great Mismatch.”
The Long Road To The Great Mismatch
Well, it might be happening. The Great Flow of Funds, every single year since 2008, into index products like ETFs (drawn, in every single year but one, from actively managed equity strategies) might be approaching its final chapter. The chart at right, repeated from last quarter’s review, shows the $1 trillion procession of assets that brought ETFs from about 3% of public stock market value in the U.S. in 2007 to 7.5% in 2014 and, more importantly – and startlingly – from about 23% of daily trading volume to almost 27%.
Chilton Capital's REIT Composite was up 6.1% last month, compared to the MSCI U.S. REIT Index, which gained 4.4%. Year to date, Chilton is up 6.3% net and 6.5% gross, compared to the index's 8.8% return. The firm met virtually with almost 40 real estate investment trusts last month and released the highlights of those Read More
Thus was created the ETF Divide, the barrier between those companies included in the major ETF baskets and everything else. Those on the inclusion list receive an automatic daily bid and higher valuations; those not on the list languish on the other side of the divide. This dynamic has been accelerating: the all-time record share of ETF trading volume of 37% was set this August 24th, when the prices of more than a few ETFs departed markedly from their net asset values. This is not supposed to happen, but at least it wasn’t boring. Shares of DVY, for example, the iShares Select Dividend ETF, temporarily dropped 35% that morning. However, the NAV of the fund declined by no more than 2.5%, and only 8 of its 99 holdings declined by more than 15%. DVY is no marginal fund; it has $13 billion of assets under management, and its holdings include the likes of McDonald’s and General Electric. It’s a sign.
This 7-year one-way flow of funds commenced in 2008, the first time equity mutual fund flows turned negative against a surge of ETF inflows. When this procession ends, the distortions upward (high valuations and artificially low volatility) will reverse, accompanied by a contraction in valuation multiples and higher volatility. The favored will come into disfavor. (Bob Dylan couldn’t have been thinking of the stock market in 1963 when he penned the final lines of that particular song: “And the first one now / Will later be last / For the times they are a-changin’.”) One can’t know, of course, in any absolute way; there can only be signs.
Not to belabor past observations, but rather, to be mindful of the foundations for our here and now, three major stimulative factors in U.S. corporate profit growth of the last 30+ years will not be repeated: interest rates cannot repeat their 13% point decline from 1981 (nor their boost to the valuations of real estate, venture capital, stocks and gold); the 25-year mercantile surge by America’s blue chip consumer products manufacturers into formerly geopolitcally unavailable markets such as China (Communism’s dying gift), beginning with the 1990 Shenzen McDonald’s, has spent itself; and the 40-year+ era of broadly stimulative government deficit spending – at least at the accustomed rate – is over. How, then, can the S&P 500 earnings grow at the historical corporate profit growth rate of the past 30 years, which – just to keep it real – was all of 4.7%, anyway?
Horizon Kinetics – The Great Mismatch
This simply suggests that one should not expect much broad-based sales and earnings growth from the large companies that represent ‘the market’. Why confine the discussion to the large, liquid companies? Because the sheer volume of funds that have flowed into index products like ETFs is simply too great to be accomodated by any but the largest. And that creates unwanted and unintended distortions. We illustrated an example last quarter with the huge mismatch between the volume of capital directed from the U.S. to an emerging market like India and the local market’s limited capacity. Only 2% or so of the rich breadth of Indian companies – of the more than 4,000 – are liquid enough to accommodate the $14 billion of AUM represented by merely 4 ETFs with the largest India investments. Result: Americans investing in ETFs comprised of very large – not emerging – companies that do not actually provide exposure to India (the majority of their sales being external) or that are excessively priced.
For this quarter’s review, an example that should make even more obvious the point that a marketplace cannot escape the reality of supply versus demand, that too much demand for even the very best idea will invalidate it. Then, we’ll get to the multiplying signs of excess in our markets.
A Short Sobriety Test
Here’s a valuation sobriety test – but we’re going to avoid equities, because the value of almost any stock, even one with a P/E of 100, may be legitimately argued. There are too many variables; who is to say what future sales and profit margins might ultimately be? A bond, on the other hand, will be worth 100 (at most) by a given date, and the balance sheet and cash flow data are pretty suggestive of the likelihood of repayment.
So, given that a 10-year U.S. Treasury Note now sells for a 2% yield, is it reasonable that, for the extra credit risk, a 10-year AA- corporate bond like IBM trades at a 3.4% yield to maturity? Most would probably say that the extra percentage point is reasonable. And for a 10-year non-investment grade bond, but from a profitable CCC+ company like Wendy’s, would another few points, for 6.3%, be reasonable? As a check, the iShares High Yield Corporate Bond ETF (HYG) yields 6.6%.
So, the sobriety test will consist of major holdings in the iShares Emerging Markets High Yield Bond ETF (ticker EMHY), and the question for each is: what should the yield to maturity be? Essentially, what price for the extra risk, bearing in mind that the best one can do is to recoup 100¢ on the dollar? Answers will be provided at the end of the test. As a frame of reference, the weighted average maturity of EMHY is 9 years, comparable to the examples above, with a weighted average yield to maturity of 8.6%. These are but two of many similar examples.
First up, 7.5% Russian Federation bonds due March 2030. Along with the sharp decline in oil and gas prices as well as gold, the revenues from which the government needs to balance its budget (which it cannot do), it fired 10,000 government workers this summer to save money. What is the yield of this bond?
Next, the Lebanese Republic 8.25% bonds due April 2021. Guess the yield to maturity. In fairness, it is not an easy country to analyze. The last year for which GDP information is posted on the website of Lebanon’s U.S. embassy is 2008. Hezbollah, which functions as a state within a state in the south, is a participant in the Syrian Civil War from the Lebanese side. The Lebanese Republic U.S. Dollar Bonds trade on the Beirut Stock Exchange, but not on most days, for reasons that should be obvious. Lebanon is a nation that could be in civil war at any minute.
Before the answers, some additional data. Russian Federation bonds have the largest weight in EMHY, at 3.6%, and Russian Federation credits total 15.3% of the fund. Lebanon is one of the top 10 allocations, at 2.7%.
How is it possible that Russian Federation 15-year bonds trade at the yield of 10-year IBM bonds? And, really, how does a nation the size of Vermont, on the brink of collapse, cowering in that particular neighbor-hood, borrow more cheaply than Wendy’s?
Would anyone seriously argue that these yields are adequate compensation for the risk assumed? If not, do the prices result from some other factor, such as artificial supply-and-demand pressures? In EMHY, new money is allocated based on float. In other words, the more debt a nation issues, the greater the allocation to its bonds because it has a greater capitalization. That is the mathematical model, and that is entirely logical – to a point.
But somewhere, a robo-advisor has just instructed someone, and an asset allocation committee for a public pension fund has just adjusted its asset allocation model, and both have decided to establish or add to their emerging markets high yield segment. EMHY has $238 million of AUM. If this one pension fund is $10 billion (which wouldn’t even make the list of the largest 300 global pension funds) and wished to allocate merely ½ of 1% of its portfolio to EMHY, that would be $50 million, or 20% of the ETF. That’s a lot. If the ETF could exercise judgment, perhaps it would allocate that $50 million other than according to the float-based weightings and other than promptly. But the mathematical model has become the reality – not a good idea. The computer is not permitted to calculate, however realistic that judgment might be, the probability of default, nor is there a “valuation” factor, extreme or otherwise, in its program – those variables simply don’t exist. Accordingly, the computer purchases additional Lebanese bonds in the precisely correct ratio. Moreover, if Lebanon issues more bonds in order to stay afloat, their total capitalization increases, and the ETF will assign a yet higher weight to Lebanon and purchase proportionately more. That is how it has a 5.6% yield.
More important than the going-in 5.6% yield, though, is the going-out yield: how will an owner of EMHY get out? What events could occur that might induce that putative robo-adviser or $10 billion pension fund to sell? And under those circumstances, to whom will EMHY sell those Lebanon bonds? And even if there are sufficient buyers, what yield will they really offer? Money came into this ETF, but can it come out whole?
Now, for the real point. Instead of thinking of this exercise as being about how indexation has so clearly inflated the prices of weird or esoteric bonds, think about it in terms of the more familiar: domestic high-yield bond funds and stock indexes, with their large weightings in Facebook, Amazon, Netflix, Biogen, American Tower, Tesla and, let us not forget, Shake Shack. That is where most of the allocations are, and it is fair to think that they are at risk.
Back to The Great Mismatch
Which brings us back to pondering the strange phenomenon of DVY on August 24th. Let’s take the most benign of ETFs, and the very first: the SPDR S&P 500 ETF (SPY). It was established in 1993, and now has $168 billion of AUM. It is a basic building block in the indexation model of asset allocation, the idea being that if long-term participation in the economic returns of the larger U.S. companies is appropriate for a someone’s portfolio, and if, say, 35% is the appropriate allocation, then that position should be held for a long period of time. One needn’t try to second-guess interest rates, economic cycles, or presidential cycles. Absent changes in one’s planning assumptions (such as retirement age), this position might be rebalanced periodically, say quarterly or annually, if it rises or falls too much. That’s the proposition of indexation – participating, not exceeding, not trading.
So, how much trading actually takes place in the S&P 500 companies? The annual share turnover rate of the largest 10 companies averages 115%; that’s 0.5% per day. Armed with that information, what do you think is the daily turnover of SPY itself, which is the index product that one neither intends to trade frequently, nor needs to? Poor earnings or a price decline in an IBM, for instance, might be more than offset by good news in Apple. They’re in the same basket. That’s the point of employing an index.
See full PDF below.