Sunday, April 27, 2014

Stock splits, odd lots, and liquidity premia

Earlier this week Apple announced a 7:1 stock split. Each share of Apple, currently trading for around $570, will be worth about $80 after the split goes into effect. Stock split announcements like Apple's are becoming ever more rare, and for good reason. All things staying the same, a stock split could typically be trusted to rejuvenate the size of a share's liquidity premium, and therefore increase the real value of the firm's stock. Thanks to changes in market structure, this stock split liquidity effect no longer exists.

The phenomena of a stock trading in the high triple digits, let alone quadruple digits (Priceline currently trades at $1,150) is a relatively new one for markets. During the decades before this one, a firm would typically announce a stock split once its shares had passed the $100 mark. Anyone reading the finance textbooks of the day—which for the most part taught that splits are irrelevant—would find this constant splitting and re-splitting to be puzzling. Academic finance, however, has tended to omit liquidity which, when combined with odd-lot trade pricing, helped drive these split dynamics.

I first ran into odd lot pricing when, as an 18-year old stock market rookie, I tried to buy 90 shares of some company, the name of which I've long forgotten. My broker told me that I'd have to pay more commission on that 90 shares than if I bought an even 100 shares. Was I fine with that? Cost-conscious individual that I am, I ended up buying 100 shares. What I had learnt was that round lot purchases, any multiple of 100 shares, would be docked at the regular commission rate. Purchasing odd lots, anything between 1 and 99 shares, attracted higher commissions, and therefore higher per share fees. After that I never tried to buy another odd lot.

This creates some interesting dynamics. When a stock has a very high price, odd lot pricing dissuades retail investors from participating in that market. As long as a share (say Microsoft) trades at just $50, small investors can easily bear the $5,000 cost of a round lot of 100 shares. But once Microsoft hits $100, the $10,000 cost of 100 shares starts to get expensive. Buying an odd lot of 50 Microsoft shares is an affordable option, but if odd lot commission rates apply, this significantly reduces the expected profitability of the position. Better for investors to hold off on purchasing $100 Microsoft and find a cheaper stock, say $50 IBM, that allows them to buy an affordable round lot and thereby avoid odd lot commissions.

Odd lot pricing therefore has the effect of pushing retail investor participation towards stocks priced in the low to mid range of the price spectrum. With a larger proportion of buyers and sellers participating in low and mid-priced shares, the ability to trade in and out of those stocks is augmented while the ability to enter and exit high-priced stocks is diminished. Liquidity is a very useful tool to cope with uncertainty—people will typically pay a premium to own it. Thanks to odd lot pricing, liquidity premia will be larger among lower priced stocks than among stocks near the high end of the price spectrum.

Having a large liquidity premium can be beneficial. Earnings staying constant, the wider a firm's liquidity premium is the higher its stock price will be. A richer stock price in turn lowers the firm's cost of capital. It can now issue the same amount of stock at higher prices, thus funding a larger project than would otherwise be the case if the liquidity premium had not been augmented.

The urge to defend liquidity premia explains why through most of stock market history, whenever a stock approached the edge of round-lot affordability, it quickly split. The higher it rose the more its liquidity premium shrunk as its base of retail investors, eager to avoid odd lot commissions, fled to lower-priced stock. By splitting its shares so that they were once again affordable, a firm could re-tempt retail investors into trading the firm's shares, returning it's thinning liquidity premium to historical levels and increasing the real value of each share. Because splitting is costless, a firm that announced a split effectively made its shareholders better off without sacrificing any resources.

Nowadays odd lot commissions are a thing of the past. Historically, odd lot transactions were routed to specialized brokers who made their best efforts to match purchases and sales, taking an eight of a dollar cut for their efforts. Later on, odd lot transactions began to electronically processed on the very same books as round lot orders. By 1991 the New York Stock Exchange's odd lot differential was removed, thanks to the ease with which computers could match odd lot tickets. Discount brokerages now apply flat pricing so that it makes no difference if a retail client purchases 100 or 99 shares.

The death of odd lot pricing has made it affordable for retail investors to invest in high-priced stock. Companies can now let their shares rise up into the high triple or quadruple digits without facing shrinking liquidity premia. I think that this helps explain the fall in stock splits over the last decade (see chart below).  In the 1980s and 90s, it was typical for sixty or so splits to be announced each year.  In 2013, only fourteen stocks in the S&P 500 announced splits, despite the fact that it was an excellent year for equity prices. So far in 2014 only four splits have been announced.

Souce: WSJ

Odd lot pricing's demise also explains the rise in the average price of an S&P 500 stock. The average price has fluctuated between $30 and $50 since 1980. Nowadays it is hitting $70. As firms are slowly discovering, modern stocks stay liquid well into the $1000 range.

Source: WSJ

But there still exists an upward limit to retail affordability. Since stock cannot be purchased in lots less than one share, as the price of a share rises towards $10,000 or so it will once again become difficult for retail investors to participate in that market. Priceline, which at $1100 is the highest price share in the S&P 500, has a long way to go before it hits that level. Until then, expect average share prices to continue rising, and share splits to remain far more muted than in previous decades.


  1. Hey JP do you think there's a psychological component to pricing? I.e. do you think that Apple after a 7:1 split @ $80 is more attractive than apple after a 3:1 split $160? Or conversely I remember how proud some investors were that they could afford shares of Berkshire class A stocks.

    I guess my question is what effect do you think nominal pricing has on-- price?

    1. I wouldn't be surprised if there's also a behavioral component to splitting. Here's an interesting paper that explores these ideas (you'll note that the authors pan the odd lot liquidity premium theory):

  2. Thanks for taking the time to share the link, great article. I like the one word summary at the end!

  3. The best case for rationalizing the argument may be BRK and BRK.B with the latter only differing in voting privilege. Another question of course is about share pricing in the UK where shares are typically prices in pence. and thus in the traditional US penny stock zone. Comments or thoughts?

    1. Nick, your Berkshire point is a good one and in preparing for this post, I did pull BRK.A and B data to see the relationship between their two prices. In general BRK.A trades at premium, but the fact that A shares have almost five times the voting power might explain this. If Buffett had designed the two shares to have equal voting rights, my guess is we'd see a persistent liquidity premium with the B's trading above the A's.

      I don't know much about stock split tendencies in the UK, but my understanding is that the tradition varies by nation. Apparently Japanese stocks almost never split.