Buybacks vs. Dividends - McKinsey weighs in

This week the McKinsey Global Institute published a study indicating that most stock buyback programs are value destroyers, in that they tend to buy high and not low.  Their specific example is a technology company that repurchased increasing amounts of shares during the stock market boom into 2008, but then stopped buying shares, even as they plunged during 2008 and 2009.

 

This seems a fitting academic underline of my point about why dividends are better than buybacks: management rarely buys shares at a significant discount to intrinsic value, whereas investors who receive dividends and reinvest them benefit from dollar cost averaging and can also choose to make opportunistic purchases (without an automatic reinvestment plan).

Management faces thre, e hurdles that almost guarantee that it will make poorly timed purchases of its own stock:

·         Management rarely has lots of spare cash when the stock is cheap.  Buybacks require significant free cash flows.  When a company is generating significant free cash flow, however, the market often places a premium on the price of the stock.  The corollary to this is that when markets are weak (and buybacks represent the best value) management usually has better priorities for its funds, including: preserving cash to ensure adequate liquidity in the face of difficult credit conditions, opportunistic acquisitions – everyone else’s stock is cheap, too, after all, and reinvesting in operations when goods, materials and labor are readily available.

·         Timing of buybacks often coincides with option exercises, which are more likely to occur when the price is high.  If the company makes corresponding open market purchases to offset dilution, the company is allowing sellers to pick the timing of the transaction, which is rarely going to work in the buyer’s favor.  Stock buybacks at nearly all firms are partially aimed at eliminating the dilutive effects on EPS that stem from new (or treasury) shares being issued as part of employee compensation.  Naturally, those employees want to exercise their options at a high price, and since multi-year fixed value options offer the holder significant discretion in when to exercise, options are more likely to be converted at cycle highs.

This is one of the most important misalignments of shareholders and management – the classic agency problem options were supposed to fix. When the same management that is selling high for its own account on the one hand is also directing the company’s cash to buy at the same time must be committing the company (and the other shareholders) to buying high.

·         Large scale buying by one investor tends to drive prices higher.  The kind of buyback activity that drives EPS and sometimes share prices higher requires significant acquisition of shares – at least a few percentage points of shares outstanding.  Such significant buying is best done by those who do not have to report their purchases, and who do not move markets.  Companies that pile up their repurchases (say in the run up to the end of the quarter) can become major market participants buying large quantities of shares in bulk and bidding up prices.  Even if management tries to buy opportunistically when share prices are weak, their own buying may eliminate the weakness.  This is good for shareholders in general – ideally share prices would always trade near intrinsic value so that all shareholders would get paid full value when they sold – but it inhibits management’s ability to create value for loyal shareholders.

Given these limitations, I generally agree with McKinsey that managements that want to repurchase stock should do so using a method calculated to minimize the disruption of the markets.   Moreover being sellers themselves they have a fiduciary responsibility not try to steer the purchases of those to whom they are selling – this is sitting on both sides of the table.

Instead, management that wants to repurchase say 25 million shares in the course of a year should simply buy 100000 shares per day on each of the 250 days or so the market is open and trading normally.  If these can be purchased in blocks near market price, fine, if they have to be purchased in lots, also fine.  If this means that the shareholders are not big enough buyers on days when management is exercising lots of options (which expire on the same day) so much the better.  This may lower the spread on the option – so what?

Finally, this raises a question about whether options should be used for compensation at all.  This is another topic, for another post.  Suffice it to say, I am not a fan of incentive programs that turns management into sellers of shares.