S&P Buyback Index

Wondering where to find a list of companies doing the heaviest share buybacks? Just go here where you can see the 100 current constituents of the new S&P 500 Buyback Index.

And you can download their list into your own spreadsheet
And you can download their list into your own spreadsheet

In fact, even if you haven’t been in search of such a list, I recommend you follow the link anyway and check out the historical performance, YTD and trailing years. Chances are, it’s much better than yours!

S&P 500 Buyback Index Performance
The 500 vs. the 500 Buyback

Why hasn’t there been more buzz about it?! Perhaps I’m not reading the right journals. I’ve tried to poke holes in the outperformance and explain it away in some fashion. Here are some avenues I’ve explored.

1. It’s just an artifact of data-mining
I mean, those trailing 5-yr returns are the result of backtesting, since the index was only formally announced in May (I assume they wouldn’t release something tainted with survivorship bias?). The easiest way to tell if something is a non-persistent data-mining anomaly is if why it works doesn’t make sense! Like stock tickers starting with B outperforming tickers starting with C, or some other nonsense. But hey, the companies are heavily buying back stock. It’s a tailwind. Makes sense.

2. It’s just high β
I put the ticker list of the current constituents into my cruncher that tells me portfolio β as a weighted average of the individual asset β’s (pulling them from Yahoo Finance). The result? 1.15 Ha! I’m onto you, buddy! – or so I thought. But if you examine June 2008 to March 2009, when the S&P 500 dropped by half, you’ll see the Buyback Index also dropped by half. So, β > 1 when the market is going up, β = 1 when the market is going down. Does your portfolio do that? Neither does mine.

3. It got lucky overweighting what happened to outperform
Maybe the index overweights what is theoretically supposed to outperform (Fama & French’s small caps and low P/B). Or maybe it’s taken big bets on particular industry sectors that have happened to work out well. These are the low-hanging fruit of explaining returns beyond what’s predicted by β.

So I went to Kenneth French’s website. I downloaded the 3-factor returns to control for value and size. I ran the multiple regression for 2013 YTD returns so far. The result?

Lower-than-predicted Beta, Minuscule Size and Value Coefficients
Lower-than-predicted Beta. Minuscule Size and Value Coefficients

As for sector exposure, I took a less rigorous approach. I randomly picked 30 of the 100 tickers and looked up their corresponding (Morningstar) sectors.

    6 of 30 were Consumer Cyclicals
    6 of 30 were Financials
    6 of 30 were Healthcare
    4 of 30 were Consumer Defensive
    3 of 30 were Industrials
    3 of 30 were Technology
    1 of 30 was Basic Materials
    1 of 30 was Energy
    0 of 30 were Utilities

It’s hard to compare this to the S&P 500 weightings, as it varies over time. But I don’t think this is terribly different (unlike, say, my personal portfolio, which is 75% consumer products).

4. It’s due to a couple lucky homeruns
In addition to tracking my portfolio excess return vs. the S&P 500, I also track what I call accuracy. Accuracy is the % of positions beating the market.

The ideal portfolio manager will have both high excess return and high accuracy. She beats the market, and it’s because her picks are consistently good. We’d all like 90% accuracy, but the word on the street is that you’re doing well if Accuracy = 60%.

Now, we know that there are certain value strategies that people have known about for decades, but that continue to work today. One version is to put together a portfolio each year of ugly, diseased stocks that no one wants. Historically, you’ve beaten the market doing this as long as your holding period is at least a decade. The catch? Bruce Greenwald says that two-thirds of the positions will go bankrupt in a typical year, or do nearly as bad. But the ones that do well, do so well they more than compensate. In theory, great. For a client, you’ve got to be kidding me. Accuracy = 33%

As for me, the bulk of my assets have been equally divided across 20 companies for all of 2013. These were hand-picked, good businesses I might never sell. I’m slightly beating the S&P so far – and you could argue I’m doing very well since Yahoo puts my portfolio β at 0.7. However, my goal is to outperform on an absolute (not risk-adjusted) basis, and only 10 of my 20 picks are currently doing so. Accuracy = 50%

How about the Buyback Index YTD? Remember: 100 postions, equally-weighted, rebalanced quarterly…

buyback_index_excess_returns

Accuracy = 75%. Just unbelievable.