After Apple And Tesla, Berkshire And Amazon Should Split Their Shares, Too
Apple and Tesla executed stock-splits last month, and their shares soared in value. This double starburst of “alpha” added an astounding $770 Bn to the companies’ market capitalizations – equal to about 2% of the entire value of the U.S. stock market. From just two companies. In less than one month.
A gain after a stock-split is to be expected (although the size of the effect here is extraordinary). A positive price response is typical, as academics have shown in many studies over the past several decades, reporting average “abnormal returns” (“alpha”) of 4% to 9% in the following year. (See my previous column, linked here, for specifics of these studies.)
The more prominent the company, apparently, the larger the effect. In 2015, Netflix – another headline stock – split 7-for-1. Its price jumped 20% right away, and was up 30% within weeks.
Stock-splits thus seem to be a reliable source of market-beating returns. Moreover, the gains are available to investors who have no prior knowledge of the event. In the Netflix case, the public’s investment window was wide open for almost a month prior to the execution of the split.
Is There a “High Share Price” Discount?
One way to interpret this phenomenon is that a share price that is “too high” imposes a discount on a company’s market value. The size of the discount would likely scale up from some lower threshold (below which there is no discount). Historical experience would probably locate this threshold somewhere between $100 and $200 a share. If this is the case, then companies with very elevated share prices like Amazon, Google and others, are actually penalizing their current shareholders by refusing to split the shares.
Activist investors today often lobby for companies to break up, spin-off business units, “de-conglomerate,” to unlock the hidden value in an over-diversified corporate portfolio. The “conglomerate discount” is well-attested, and ranges around 10-15% on average for diversified companies in the U.S. Is it possible that a “too-high” share price can also depress a company’s market price below its “true” value? Lobbying for a stock-split to remove this discount and unlock their value could become another interesting strategy for an activist investor.
The Strange Dearth of Stock-Splits
Notwithstanding this opportunity for companies and investors to gain market value, stock-splits are out of favor.
The drought is hard to explain. Perhaps fractional share trading is part of the answer. Brokerage firms have introduced this innovation lately, allowing small investors to buy “slices” of a single share of high-priced stocks such as Amazon (trading over $3000 recently). This may reduce the urgency of lowering the share price to attract “retail” investors. But the decline in frequency of stock-splits started before the fractionalization trend. It is likely that these companies – which are often dominated by founders (Bezos, Sergei and Larry) who are less concerned about increasing their already spectacular net worth than with maintaining control, have simply decided to let the share price run for other reasons. High share prices discourage “hot money” and day traders. Perhaps companies have also come to view the elevated price as a badge of prestige. Economists recognize that high prices are part of the brand equity for luxury goods. You expect a Rolex to be expensive. High price is taken as a sign of high quality. Perhaps the way these companies position their shares is based on a similar mindset.
In any case, the whale of the anti-split crowd is Berkshire Hathaway. Warren Buffett hates stock-splits. When he made his first purchase of Berkshire Hathaway shares in 1962, he paid $7.50 a share. He has never allowed that stock to split. Today a single share of BRK-A trades for about $330,000. Volumes are minuscule (frequently just one share per trade, and only about 1000 total shares per day). It’s an exclusive club, and they’re proud of it.
- “Were we to split the stock [writes Buffett] or take other actions focusing on stock price rather than business value, we would attract an entering class of buyers inferior to the exiting class of sellers. Would a potential one-share purchaser be better off if we split 100 for 1 so he could buy 100 shares? Those who think so and who would buy the stock because of the split or in anticipation of one would definitely downgrade the quality of our present shareholder group.”
Buffett has compared doing a stock-split to undergoing a colonoscopy – which certainly expresses his discomfort at the thought of cheapening his shares for the hoi polloi.
Nevertheless… Berkshire Hathaway did split its stock once, in 2010. Actually, twice, sort of.
The story is interesting. In the 1990’s, when BRK was already trading at $34,000 a share, too expensive for most retail investors, a Philadelphia fund manager came up with a scheme to make it possible for them to “own” the stock anyway. He proposed to create a fund called “The Berkshire Trust” which would invest in “one thing only” – BRK. The fund would sell its own shares at a more reasonable price, tailored to smaller investors. (Later the projected name was changed to “The Affordable Access Trust” — which better described its intention.) Apparently, Buffett hates “middlemen” even more than stock splitting, so to kill the idea, he grudgingly launched a new share class, BRK-B, priced initially at 1/30th the level of the existing (now BRK-A) shares, but with much reduced voting rights.
Then in 2010, Buffett faced another challenge. To pay for the acquisition of Burlington Northern railroad – which involved acquiring shares from small holders of Burlington stock by using Berkshire stock – Buffett authorized a huge 50-1 split of BRK-B. At a stroke this turned BRK-B shares into a more typical equity instrument, priced at about $68 each. (They trade today at about $220 a share.) Among other things, this gesture of conformity with market norms finally allowed Berkshire Hathaway to join the S&P 500 index.
Did the BRK-B stock-split generate the same sort of alpha that we see in the Apple, Tesla and Netflix cases?
Yes it did. Here is the raw price trend data on the 2010 split.
The weekly pattern of alpha accumulation — that is, the “abnormal” gain above the overall market trend – is revealing. The chart here shows the BRK-B gains relative to the S&P 500 average during the first quarter of 2010.
The alpha was concentrated in the first two weeks after the split.
Note two other facts. First, the 50-to-1 split was announced in November 2009. So unlike the Apple and Tesla (but similar to Netflix), it was the execution of the split rather than the announcement that drove the gains. As with Netflix, investors here had substantial prior notice of the event, ample time to absorb the information and position for a profit. And even after the split took effect, the market mispricing persisted. Even if an investor waited a full week to buy in, there was still an opportunity to profit.
As already mentioned, the other event that came into play in this period occurred on Jan 26 2010 – just five days after the split took effect: Berkshire Hathaway was added to the S&P 500 index. Its newly-split shares now traded with sufficient liquidity for the first time.
Berkshire’s addition to index may account for some of the gains shown here. Or not. Joining an index is often a mixed signal for a stock, and indeed many “joiners” underperform afterwards. Note that the Berkshire alpha goes negative after Week 2. (We’ll consider the effects of index membership changes as another source of alpha in an upcoming column.)
What About Amazon?
The Apple and Tesla splits last month have raised the question – now debated actively in the online forums – as to whether Amazon should also go for a stock-split. Until now, Jeff Bezos has been “dismissive.” But the evidence for the “high share price discount” is strong. If Apple – the most valuable company in the world today – could add more than half a trillion dollars of market value in less than a month following the announcement of a split, effectively “unlocking” 30% of its hidden value – it suggests that Amazon, as successful as the stock has been for investors in recent years, could still have untapped alpha embedded in its capital structure.
There is a direct precedent here. Amazon split its stocks three times in the late 1990s. After the first split (June 2, 1998), the stock doubled in 3 weeks. From a month before until a month after the second split (January 5, 1999), the stock doubled again. And in the month following the third split (September 2, 1999), the stock rose by another 30%.
Do stock-splits – “meaningless” as they are said to be – really cause these gains? (A split is supposed by many analysts to be a “meaningless” event in terms of the company’s value, the same as exchanging a dime for two nickels).
I would answer a partial-yes. Splits are a causal factor, but only to a limited degree – mainly through increased liquidity (more shares, higher trading volumes), which attracts a wider range of investors, both large and small. The more important effect of a stock split is as a signal of strong business performance, current and forthcoming. Companies split when their shares have been rising and they expect prices to keep on rising, driven by robust fundamentals.
I have no inside information (of course) and no sense of how Amazon might decide this matter today. But if they announce a split it will likely present another window of opportunity for ordinary investors to realize above-market gains. Perhaps even Warren Buffett will come along again at some point.