I had never paid much attention to Horizon Kinetics, an asset manager run by Murray Stahl and Steven Bregman, but I've been reading their market commentaries and research reports and they seem like a good buy and hold manager. [The problem with buy and hold, of course, is that it has such big drawdowns unless you get out of the market during periods of overvaluation that can last for ages.]
The 2012 Q4 letter [pdf] presents a rather unique idea that the metrics most investors focus on (P/E, P/B, EV/EBITDA) could in some cases be a manifestation of availability bias.
"[T]he idea of assessing a varied set of companies via one or even several metrics, as if a portfolio of companies can all fit within a handful of pre‐defined boxes, is a very limiting, unidimensional way to view businesses. It can miss a great deal of information content. What about all the data that has not been predefined and made available as individual fields in the databases that produce those ready made reports? In a sense, for people who rely on those tools, that data doesn’t even exist—and it might be very important. Availability error can be very costly"Unfortunately, they use this concept to justify owning the Beijing Airport operator, which apparently trades at half the valuation per passenger as Heathrow and the Paris airport. Although the Beijing airport is growing more quickly, the GDP/capita of England is 3x Beijing (and 6-7x China as a whole), so it's not immediately clear that this discount is unwarranted.
They do call attention to another very interesting metric, and that is the rate of share repurchases. There's a lot of academic research about the effect of net stock issuance on future returns [study pdf], and there is a clear inefficiency where the market is too slow to recognize the effect of repurchases (Conrad Industries) or new issuance (say, GMX Resources).
If I was going to buy and hold equities, I would be interested in a lot of the big share repurchasers they mention: Berkshire, Dundee, Leucadia, Liberty Media. The common themes that unite their equity holdings are: repurchases, "owner-operators" (managements with very large equity stakes, not free options), and conglomerate discounts.
Another recent letter was about companies with dormant assets [December 2012 pdf]. For example, Dish Network has several billion dollars of wireless spectrum that the market doesn't really give it credit for; Icahn Enterprises had not really been getting credit for buying the $2b Fontainebleau resort for $156mm, etc.
The big discounts for conglomerates and dormant assets reminds me of Peter Thiel's "indeterminate world" idea from last summer. As he wrote, "any company with a good secret plan will always be undervalued in a world where nobody believes in secrets and nobody believes in plans." As HK put it in an earlier 2012 letter
"If one were to draw a Venn diagram of the critical, motivating elements in the financial markets today, two of the largest intersecting circles should be labeled Ultra-Low interest Rates/Reinvestment Risk and ETF Proliferation/Equitization of Asset Classes."Think about what is really expensive in the market right now: fixed income and assets that share the salient property of fixed income: predictability. So in addition to bonds, that means utilities, REITs, MLPs, royalty trusts, apartment buildings, even single family rentals which investment managers are trying to scale for the first time (won't work).
Why is predictability so valuable? I'd argue that, just as Thiel says, investors have largely lost faith in the ability of managements to plan and execute for the benefit of shareholders. The opportunities for cash to disappear on the way to one's pocket are too numerous: poorly timed buybacks or capital expenditures, dumb acquisitions, obscene salaries and options grants.
So public markets put a huge premium on predictable, non-divertable cash flows and more attractive (lower) multiples on cash flows that are seemingly uncertain. As for conglomerates, nothing gets punished like moving parts. A bird in the hand is worth three in the bush.
[Obviously, there are various exceptions to the idea that investors don't believe in plans. For example, investors clearly believe in the Amazon plan. But it's clearly the case that there is a predictability premium.]
Think about this in relation to Conrad Industries. We know that shipyards are essentially booked for this year and next year, and that we are finally embarking on a number of years of steady GOM drilling again after the ban. Conrad will make at least $35mm EBITDA in a good year. The current enterprise valuation, net of excess cash and the probable BP settlement, is about equal to two years' EBITDA. It's as if the company gets no credit for the future years' earnings that aren't already known.
Hunger for predictability is also why the arb opportunity for David Einhorn's opportunistic use of preferreds exists. That hunger has created a wide disparity between the cost of capital of debt or preferred stock and the existing common stock of most companies.
A Conrad Industries preferred stock would trade at a single digit yield; say 6 percent. The earnings and cash flow yield on the common stock is much higher than that at the current share price. [For Conrad, it's something well north of 20 percent depending on whether you look at net income, EBITDA, or FCF; and depending on how much excess cash is on the balance sheet.] If the company issued preferred stock and used the proceeds to buy back common, it could capture that spread. So $20 million would mean several million dollars in additional annual income to common shareholders, assuming that many shares could be bought back near the current trading price.
Another arb that it creates is what I call the predictability arb. Chesapeake has thought of this because they mentioned it in an investor presentation.
"recognize the world is 'short yield' and [want to] provide various investment opportunities to yield-hungry investors who are willing to pay more for select assets than core E&P investors are willing to do."It looks like E&Ps, like Chesapeake, can be thought of as a bundle of two dissimilar types of assets: producing properties that income investors would pay a huge multiple for, and undeveloped resources that investors are suspicious of because no one believes in plans. Rightly or wrongly, the claims of high potential IRRs on new wells seem to fall on deaf ears.
The predictability arb is to develop producting wells - with the predictible cash flows that investors crave - and spin them off into trusts. With each sale, use the cash flow for drilling new wells or share buybacks if that has a higher IRR.
If the public markets are skeptical of your plans, take anything predictable and drop it down into a REIT, an MLP, a trust and spin it off. Then decide whether the best use of the proceeds is stock buybacks, capex, or acquisitions.
What I find interesting about Horizon is that Stahl and Bregman seem to have this idea, and it is the most likely to work out of the ways that I see people allocating long-only capital. Oddball Stocks wrote a post about Investing in Horizon Kinetics through the back door with FRMO, which is an entity run by Murray Stahl and Steven Bregman that owns about 1 percent of HK.
Asset management is a fantastic business, and the end of the interest rate cycle will reorder the industry. Rising interest rates will be brutal for those who are too far out on the curve. But surviving asset management firms will be worth more in a higher interest rate environment.