- Warren Buffett, 1994 letter to shareholders.
According to Buffett:
1. A strategy of paying less than the present value of future dividends should be profitable.
2. I’m incapable of coming up with an objective estimate of present value.
It may not be immediately obvious but under certain circumstances it’s possible to profit from former while avoiding the problem of the latter.
A practical example
For a brief moment in 2009, Berkshire’s A-shares were worth 1575 x Berkshire’s B-shares. This means one could sell an A-share, buy 1500 B-shares and have some money to spare.
By definition, 1500 BRK.Bs have dividend rights equal to one BRK.A. Berkshire knows this and instructs a broker to do the trade until the gap closes. It’s closed. Here's a chart of the spread of BRK.A versus 1500 x BRK.B.
A current practical example
Hubbell Inc. A-shares (NYSE:HUB.A) closed at 44 while the B shares (HUB.B) closed at 50.
The A shares have volume of a few thousand per day. For the long side of the trade, one should be able to build a position of $ 10k without moving the market.
The (expensive) B shares are more liquid with volume of 200k shares or more. Here's a chart of the spread of HUB.A versus HUB.B. You'll notice the A shares were priced at a double-digit premium in 2008. Today they trade at a significant discount.
Dual-class shares are usually created to guarantee control for founding family members or other insiders with a minority stake in the company’s cash flows. Corporate charters typically require dividends and distributions to be equal or proportionate for all classes of shares.
Whether or not the separation of control and economic interest is a good idea is of no importance to the investor putting on a pair trade. In fact, if the business is run into the ground, all equity is wiped out and the trade is highly profitable.
In any case, voting rights have zero value if we stick to the definition according to Warren Buffet (and John Burr Williams before him).
The nice thing about these opportunities is that the risk is within the span of control of the investor. The usual risks (business, fraud, storm, earthquake, terrorists, bear market etc. ) are cancelled out. Anything that causes a stock to drop is in fact good for your wealth. The reduced value of the short position more than compensates for the drop of the (twin) long.
There is some market risk. A bull run or short squeeze may cause the margin requirement on the short position to force you out at the worst possible time. I suggest limiting each short position to 5% of your portfolio.
If the run is caused by someone bidding for the company (or rumors thereof) the gap closes in the blink of an eye and you simply exit the position at a profit.
In practice, if you're squeezed out, you are seeing your short positions go up sixfold while your longs stay put. If that happens, consider closing all your positions. Permanently.
Let’s see how this works.
1. We short 1000 B shares of company X at $10.
2. We simultaneously buy 1000 A shares at $ 8.
We assume the cost of this position is 500 per annum (all included).
Assuming the spread closes in 2 years, we gain 10 000 – 8 000 – 2x500 = $ 1000
You'll notice the calculation of returns is sensitive to the length of time it takes to close. Historically, 80% of spreads over 10% close in less than a year. Read more: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1338885
There are two additional and unknown factors.
1. Your broker will require some collateral to put on the short (margin).
2. You immediately generate $ 2000 of investable cash. Depending on your skill, this will generate some return in the two years it takes for the gap to close.
In a sense it works like leverage. It has two advantages though:
1. If everything suddenly goes south, your pairs will be profitable. At that point you will be in a position to invest more . This is not usually the case if you’re levered.
2. The extra investable cash is free of interest.
This works like the float of an insurance company. If you’re disciplined, you get to invest free money.
There may be dozens of publicly traded dual-class shares with price differentials that clearly do not reflect intrinsic value. With some work, I could construct a monthly list of pairs and include the spread and cost to borrow. We would be looking for high spreads in combination with low cost, see how this works in practice. The list would consist of stocks that:
1. Have a spread of at least 10 %
2. Have enough liquidity to short at least $ 1000 worth of shares at reasonable cost.
3. Trade on the NYSE, Nasdaq, TSE, LSE or Euronext
There's work involved so I need to know, would you be interested ?