Seeking Alpha contributor Macro Investor recently wrote a follow up article, More Proof That Share Buybacks Do Not Add Value, to his original piece on the subject, Smart Investors Should Ignore The Gimmick Of Intel Share Buybacks. In his first article, he took Intel (INTC) management to task stating that Intel's recent debt offering to be used for share buybacks was a "gimmick" that "smart investors should ignore" the debt-financed buyback because it doesn't add value to the company. In his follow up article, he posits that because the PowerShares Buyback Achievers Portfolio ETF's (PKW) massive outperformance of the major stock indices over the past five years is somehow due to the Federal Reserve's quantitative easing programs and is not "statistically significant," it too should be ignored. This is used as "proof" that buybacks do not matter and, again, should be ignored by investors. I will use this article to show that this hypothesis is imprudent.
I posted a rebuttal to Macro Investor's first Intel article titled, Intel's Buybacks Are Not A Gimmick: A Rebuttal, in which I postulated that share buybacks alone are not a reason to buy a stock but can be a great reason to stay in the stock. This is due to the fact that, by simple math, if a company is decreasing share count, each dollar of earnings is attributable to fewer shares, which makes earnings per share, or EPS, rise. In conjunction with that, rising EPS is usually accompanied by a rising stock price as the price to earnings multiple must readjust to the lower share count (and higher EPS). Examples are given that show what could potentially have happened to Intel's share price without the buybacks as well. I won't rehash my entire article here, but this premise is once again rejected by Macro Investor, and I'd like to try and set the record straight.
PKW Versus SPY
First, Macro Investor uses a five year graph of the performance of PKW versus the S&P 500 proxy ETF, (SPY), to show that PKW has indeed massively outperformed the broader market over the past half-decade (graph is below, again this is his graph, not mine). However, the author then goes on to explain away this outperformance as having to do entirely with the Federal Reserve's quantitative easing efforts, starting in 2010.
Now, I may not be as smart as Macro Investor but I have absolutely no idea what quantitative easing has to do with companies that perform share buybacks outperforming major stock indices. Seems to me, the Fed's QE programs aren't targeted at companies that perform buybacks, or any other subset of companies for that matter (except, perhaps, at banks). This argument makes absolutely no sense and the author doesn't really provide any backup of this theory, so I will just move past it. He seems quite sure this is the reason for PKW outperforming, but doesn't impart any of his reasoning in the piece, so I am forced to assume there isn't any.
Next, I didn't see any components of PKW that aren't also in the SPY on the fund's website. Therefore, it seems to me that the outsized performance of PKW over the SPY over the past five years would actually make the outperformance less pronounced, as the members of PKW have actually served to raise the SPY's performance over the same time period. Without extensive research, it would be difficult to quantifiably show how much PKW has helped the SPY over the past five years but given the gigantic spread between the two ETF's performances, it can be reasonably assumed that the "help" PKW provided to the SPY is significant.
Indeed, if we look at the holdings of PKW, we can see that some of the largest companies in the SPY are included. This means that comparing PKW and SPY's performance and concluding they aren't that different is akin to comparing Apple (AAPL) to the Nasdaq (QQQ); of course their charts are going to look similar. This argument, again, does not hold water with me as I believe the author is using a premise that should not be accepted.
Given my points about comparing SPY to PKW above, I think a better comparison is the iShares Russell 2000 Index Fund, (IWM). I didn't see any PKW components that are also part of the IWM; even if there was one or two, the overlap should be so insignificant it can be reasonably ignored. What we see in comparing the PKW to IWM (and SPY) over the same five year time period is, again, massive outperformance on the part of PKW. Note: this graph is mine, courtesy of YCharts.
Without the overlap of companies that are in both the SPY and PKW, we see very similar results. That is, the companies that have performed buybacks outperformed those companies that can be assumed not to have performed such extensive buybacks, IWM components. The reason I chose the Russell 2000 is because it is comprised of small companies that should be in the growth stages of their existence. The point I'm making here is that the Buyback Achievers still tremendously outperformed even small, growing companies.
As Macro Investor correctly points out, during the financial crisis, PKW and SPY looked very similar. However, in 2010, when the large divergence started to occur, Macro Investor purportedly thinks this is somehow due to quantitative easing. I disagree and think instead that this outperformance occurred because earnings power that had been decimated for many companies during the sell-first-ask-questions-later meltdown of the financial crisis was reverting back to the mean.
In other words, when earnings started to rebound for companies performing significant buybacks, the EPS component was growing more quickly than it otherwise would have because of the reduced share counts, a result of buyback programs. The companies that are in the PKW were growing earnings, as were many other companies, but since they were also decreasing the share count through buybacks, their EPS component was rising more quickly than the broader market. This resulted in those companies being assigned higher prices because their earnings per share were growing more quickly than the broader market. The results of this reversion are seen below:
Lastly, I'd like to discuss Macro Investor's conclusion about the extensive statistical work he performed for his articles that he believes proves buybacks are meaningless. He says:
"R2 is 92%, which means this ETF basically tracks whatever S&P 500 does. The intercept is 0.14%, with 95% confidence intervals of -0.2% to 0.46%. This means that statistically speaking, there is no proof whatsoever that the real value of the excess monthly return from this ETF over that of S&P 500 is positive. In fact, it could be negative as well."
First, he put a lot of work into this statistical analysis and should be applauded for his effort. What I believe, however, is that while his methods are sound, his interpretation of the results is illogical. First, he states that because the R2 value is 92%, the ETFs are essentially the same. This is largely specious in my opinion as you could randomly select any 20, 30, 50 or 100 stocks over $100M in market capitalization and I'd be willing to bet the R2 for that basket of stocks would be north of 90% to the SPY as well.
Second, the intercept of his regression line is 0.14%. Again, I'm not statistician but I can interpret this to mean that since the data points are monthly returns for the PKW versus the SPY, the PKW enjoys a 0.14% "head start" of outperformance each month. Annualized, this is roughly 1.68% of outperformance.
Third, the 95% confidence intervals produced by his thorough analysis are -0.2% to 0.46%. The author then says that, "This means that statistically speaking, there is no proof whatsoever that the real value of the excess monthly return from this ETF over that of S&P 500 is positive. In fact, it could be negative as well." I don't know about you, but if I could have relative performance of between -0.2% and 0.46% every month, I would take those odds every time.
I believe Macro Investor is missing the point that, given that confidence range, it means PKW is outperforming SPY 70% of the time that his analysis covers, assuming random distributions of returns. Of course, since the confidence intervals only cover 95% of monthly returns, we can conclude that PKW outperformance versus SPY would be as high as 70% but possibly as low as 66.5%. Regardless, we are talking about a two in three chance of outperformance each month.
The point of all of this analysis is that buybacks can be great value contributors if performed properly. We have all heard of companies that botched buybacks and wasted billions of dollars of shareholder money. However, when executed properly, buybacks can indeed provide a boost to EPS and share prices. Buybacks are not always the right move for corporate managements, but they should not be disregarded completely just because of what they are.