Time to Split!!

As many of you are probably aware, Berkshire Hathaway has announced its intent to split its class B shares and the requisite "deep analysis" of whatever Buffet is doing has journalists chasing the story.
http://money.cnn.com/2009/11/06/markets/thebuzz/index.htm
Since Berkshire is not the first company to ever split its stock, it is worth looking at key questions that come up anytime there is a stock split.

1. Do companies split their stock often?
The answer is yes and no. Some companies are serial stock splitters, splitting their stock at regular intervals. Other companies let their stock ride. In fact, Berkshire Hathaway is a classic example of a company that has avoided splitting its shares, with it's class A shares trading at about $100,000/share.

2. Why do companies split their stock?
There are several reasons provided, though not all of them hold up to scrutiny:

a. Attract new investors to the company: There is a belief that some small investors and even a few institutional investors either cannot or will not invest in companies if the stock price rises above a threshold level. The "level" itself seems to be a malleable number and vary across companies. There is little evidence for this proposition and even if there were evidence, so what? Inherently, there is nothing good about attracting investors who have hitherto avoided buying your stock and it is entirely possible that these investors may bring with them preferences for dividends and other corporate finance policies that put them at odds with the firm's current policies.

b. Improve liquidity: This is the time honored argument provided by many companies when they split their stock. Having a lower-priced stock, they argue, will increase trading volume and improve liquidity. The evidence, though, points in the opposite direction. Aggregate trading volume does not increase significantly after stock splits and transactions costs go up (not down). The reason for the latter effect is that the bid-ask spread, as a percent of the stock price, tends to be higher for low-priced than high-priced stock. (Try a simple experiment. Try buying 100 shares of a stock trading at $200/share, 1000 shares of a stock trading at $20/share and 10000 shares of a stock trading at $2/share and figure out your total transactions costs with each, including commissions and bid-ask spreads.)

In the case of Berkshire Hathaway, the reason for the split lies in the recent acquisition of Burlington Northern. Berkshire had offered shareholders in Burlington a choice being paid in either cash or Berkshire class B shares. Since Berkshire's class B shares were trading at more than $ 3000/share, there were many small stockholders in Burlington who could not avail themselves of the stock offer. (If you owned less than $3000 worth of Burlington stock, you had to settle for cash.) This has tax consequences. When you as a stockholder in a target company accept cash on an acquisition, you have to pay taxes on capital gains immediately. If you receive shares in the acquiring company, you can defer paying capital gains taxes until you sell those shares.

3. How do stock prices react to stock splits?
Are stock splits good or bad news? There have been several studies of stock splits over the last few decades and the findings can be summarized as follows:

a. At the time of the stock split, there is, on average, a very small positive impact on prices (about 1-2%). In other words, when there is a two for one stock split on a $50 share, the new shares trade at about $25.25. This is usually attributed to a "signaling effect", where markets view stock splits as a sign that the company expects earnings or dividends to increase in future periods.

b. There is some debate about whether investors can generate higher returns in the period after the stock split. While many of the earlier studies indicated that stocks that split did not "beat the market" in the months after, more recent studies provide evidence that "stock split" stocks generate significantly higher returns.

c. As with all investments, there is another shoe waiting to drop. Studies also indicate that the volatility increases after stock splits. In a very general sense, a stock split seems to increase both returns and risk.

If you are interested in reviewing the literature, there is a good survey paper on the topic. You can get to it by going to:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1362259

The bottom line. Stock splits, for the most part, are cosmetic and should not play a central role in whether you invest in a company or not. The reason is simple. From an intrinsic value standpoint, changing the number of units you divide the value by cannot change the overall value of a business. If everyone gets the same percentage increase in units, there cannot be winners and losers within the firm. However, the evidence does suggest that they can play a secondary role in stock picking. Thus, when faced with investing between two otherwise equal companies, one of which has split its shares recently and the other not, you would go with the first one.

Two side notes. First, everything that I have said about stock splits also applies to stock dividends. In fact, stock dividends represent an even bigger pain in the neck, since they leave investors with strange share counts - 100 shares become 102 shares. Second, reverse stock splits, where a company whose stock is trading at a very low stock price offers 1 share for every four or five shares, seem to be more defensible from an economic standpoint, since the transactions cost argument works in your favor)