Subject: Re: From the letter
"As for the S&P500, after about a third of profits are sent to index shareholders as dividends, more than 100% of the retained balance is used repurchasing shares ... Share reduction of the index companies was a modest 0.7% per annum for the past decade. Said differently, companies spent roughly 60% of profits to purchase 2.7% of their market capitalization each year, yet only reduced the share count by 0.7% annually.

SteadyAim:
Is this right? Over what timeframe? I knew buybacks offset issuance to some extent, but this is saying that divis are only 33-40% of profits, and all of the rest goes to buybacks (not investment??) and only ~25% of that represents share count reduction. So shareholders are only getting say 36%+16% = 52% of profits? Is it really this bad in the long term? It doesn't seem possible for US companies / markets to do so well if this is the case.
What am I missing?

I want to congratulate SteadyAim firstly for publishing the 1,000th post above! Secondly, I congratulate for the post containing many questions, which is the core spirit of Shrewd'm. It doesn't rhyme so well, but rather than curiosity killed the cat, it should really be curiosity allowed the cat to kill.

No-one responded to your excellent questions. The main thing is to think of value given to shareholders in two main components; (1) the dividends paid out (2) plus the inherent increase in underlying value of the firm itself. Buying back stock - in itself - doesn't increase the value of the firm, and often decreases it. Buybacks are economicaly just like any other stock purchase; whether purchasing your own firm or someone else's firm, the value only increases if you get more than $1 worth for each $1 spent. Similarly, buybacks are not "returning value to shareholders" as constantly quoted and almost perceived as a fact. If Microsoft buys some stock of Google, it isn't "returning value to shareholders" just like if Microsoft buys stock of Microsoft it also isn't "returning value to shareholders". Unless buying at a bargain (or - technical note - if promising, in a moment of insanity, that the cash would otherwise be held forever and never used in any other way other than buybacks. Companies never make the later insane promise of course, and buybacks merely eliminate the future more intelligent use of the cash).

In fact, sometimes buying other firms, rather than your own firm, is a much smarter idea. Sanborn Maps is a very good example of this. In the 1930s their mapping business was dwindling, and instead of buying stock of Sanborn as would be popular today, they kept buying none of their stock and instead built an investment portfolio of outside stocks - which eventually greatly exceeded their mapping business - which Buffett could smell, and the smell was like flowers, and he quickly became excited.

The main answer to "It doesn't seem possible for US companies / markets to do so well if this is the case. What am I missing?" is capital investment. Our businesses spend our cash on stuff that shows up on the balance sheet, and (if purchased well) the additional purchases produce extra earnings. This "new stuff" (from capital expenditure) includes capital expenditure to expand current operations, and the purchase of other businesses. So value "obtained by shareholders" (a term I'm inventing on the spot, as it is exceedingly more useful than value "returned to shareholders") is the sum of dividends paid out and the "new stuff" purchased.

How much does this "new stuff" increase the value of the average, though? The answer is not as much as most would hope. After inflation, stock prices only rose 2-3% or so over the last century (if you average the starting price over a few decades, given that in 1923 stocks were exceptionally cheap), and the actual real value increase was generally closer to the 1-2%, as 1% of the stock price increase is just from the increase in valuation. We pay about double now what we used most of the last century (the CAPE ratio is about 28 now, versus fluctuating wildly around a central value of about 14 over the last century). So, with the "extra stuff" providing us only 1-2%, and dividend yield usually around 3-4%, it is the dividends paid out that account for most of the value provided to shareholders.

(NB: The dividend yield of the S&P500 today is 1.6% but if stocks were half their current value, the yield would be 3.2%. It is well known that dividend yields are lower today than a century ago, but that isn't only because payout ratios are much lower; it is also just a result of stock prices being higher now. Investing from any random starting point in the last 100 years was usually more fun than having to start from today. With the dividend yield so much higher, you were usually likely to have a better return than starting from today.)

It is true that buybacks (which decreases the share count) these days are largely negated by employee stock compensation (which increases the share count), but it doesn't matter in itself. Remember the buybacks are no different to the company having an internal stock portfolio and just buying the stock of its own firm (so it isn't "returning value"). Stock based compensation is accounted for today by reducing the reported earnings accordingly, and the stock based compensation is also recognised as a non-cash expense on the income statement and added back on the cash flow statement. When CFOs state "We are buying stock to negate for stock compensation" they are misunderstanding the economics. The stock is diluted, which is one thing and, but the decision to buy stock is a completely independent decision that should only be an equation of the immediate economic value received (hopefully more than $1, though this isn't usually the case) for each $1 spent.

- Manlobbi