Buybacks, or dividends?

Hello - So buybacks seem to be on everyone’s minds. I think it’s time for an explanation of what Wall Street has been up to.

~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~ The theory of markets

~~~~~~~~~~~~~~~~~~~~~~~

~~ Feel free to skip this segment, or skim if you understand capital allocation. The important discussion on buybacks happens after this segment.

The goal of any investor is to hit an 8% Return on investment. It’s as simple as that.

All stock prices continuously adjust and fluctuate as investors and traders try to figure out and come to a consensus on exactly what is the expected value of the return of any given company. That’s all the stock market is, is a consensus algorithm.

Any consensus algorithm could work, the more bureaucratic / government-sponsored version of a valid consensus algorithm would be one where your average person goes to school and gets a 4-year degree in “predictive returns”, and then spends their time getting paid a set $50k/yr salary making predictions on companies. This could be what would happen in a communist government for example, or in a heavy-handed government. This would cost a notable amount of money for the government to be spending $50k/yr per analyst to run this program, and the people getting paid to do this aren’t contributing to society in other ways, and they would in general likely suck at capital allocation. Additionally, since bureaucratic systems are neutral when you do well (You still make $50k/yr), and punish you when you lose (You get “repositioned”), such models of capital allocation would be far more risk averse and thus strongly suboptimal. Hedge funds also look to be risk averse, but they use options such as calls and puts and shorting of similar stocks to hedge, which is something that a bureaucrat would likely never discover because there would never be a “degree” in options until after it’s invented, and there’s no incentive to invent it because no one benefits.

By instead using a consensus algorithm such as a stock market, the people who can consistently predict which companies will hit the glorious target of 8%, will exponentially accrue more and more capital that can be reallocated according to their out-performing predictive algorithm. And the people who consistently mis-predict which companies will hit the target, will exponentially lose their capital until they exit the market and instead just invest in the S&P500 as a whole (Or, as opposed to lose their capital, simply gain capital at a slower rate and will inevitably give up and buy SPY). People who have an unbelievable ability to consistently and accurately predict which companies will generate large returns, will be able to command enormous amounts of capital, such as Warren Buffet’s Berkshire Hathaway which allocated $400B+ capital, and whose stocks seek a premium of 25% as their holdings had consistently outpaced the S&P500 (Of course, that premium will readjust the stock price of BRK.B so that BRK.B does not outperform the market, it simply hits 8% per year targets). I say "had", he hasn't done too well, but no one has. Hedge funds and capital allocators definitely get hammered during 10-year bull runs because it's impossible to ever catch up with SPY when this happens, it's a hard comparison vs 1999-2009 which showed a loss of 50% despite 10 years in the market. But, it all evens out, 10 years of 12% gains brings SPY back into 8%/year.

To be honest, you can absolutely come up with alternative capital allocation consensus algorithms that involve bureaucrats getting paid but they play the game that the capitalists do only with fake money, their pay is proportional to the money that they make, while using their decisions to push real government-sponsored investments. The only problem is that this would only work on public markets, and every company begins as a private company, which is far less structured and easy to analyze. And, the moment you give them some proportion of their winnings, you’ve just reinvented the wheel. The difference being in a bureaucratic system you would pay the bureaucrat a portion of earnings, and in the capitalist system the capitalist pays you a portion of their earnings in taxes. (~15% of capital gains, which debatably should be higher but it’s set in order to counteract the 21% corporate tax that results in double-taxation when investing in companies and remains competitive with the 36% income tax that someone of that tax bracket traditionally gets taxed, but which inevitably compels companies to reinvest instead of pay shareholders since reinvestments don’t get taxed - so eh, it gets the job done)

So - Why is this important? Just from an abstract view, you can understand a lot about a market.

For example, with the markets, if capital allocators predict that a stock will return 12% per year, they will flood capital towards that company until its stock price rises to the point that it only returns 8% per year on average. If capital allocators predict that a stock will return 5% per year, they will pull capital out of the company until its stock price calls to the point that it again returns 8% per year on average. Of course, that’s only the average, it might return -5%, +20%, -70%, +65%, and anything in-between. But to the best of the ability of investors, we all predict it will return 8% per year ad infinitum. And if anyone thinks they’re wrong, they can buy low and sell high, with their purchase being their pitch into the consensus algorithm that will push the stock price just a little bit higher as it adjusts. Of course it’s impossible to predict the future, stocks like AMZN continue to beat expectations, and stocks like SNAP have underperformed, but hindsight is 20/20, and you’d be a millionaire if you could at the time predict which ones will be overperform and which ones will underperform - the general iffy performance of Hedge Funds show that this is far more difficult than it seems. Every year AMZN appears to have topped out and finally hit it’s cap, and it yet again beats expectations, but it could stop at any moment. Poor performers sometimes wake up, like MSFT which putz’ed around for almost 15 years woke up with Azure and started to soar like a rocket; all of the future possibilities are integrated into the trade price which continues working through each day as a random walk 50% up 50% down, while difting week-by-week around the inherent probabilistic value. As always, hindsight is 20/20, and the market opens every day asking for you to make your attempt to prove that you know something that everyone else doesn’t, and if you can't figure out who will win and who will lose, than you must accept that to the best for your ability and knowledge, a given stock will return 8% per year on average at its current share price.

~~~~~~~~~ End Introduction to Capital Allocation, The Theory of Markets ~~~~~~~~~

~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~~ What's the deal with buybacks?

~~~~~~~~~~~~~~~~~~~~~~~

So - when a company decides between dividends and stock buybacks, in what ways might they affect an investor? Well, the answer is in reality, none. You can deduce this theoretically just from the above abstract understanding of the market as a whole, you cannot create money out of thin air, a company’s income is irrevocably attached to the investor’s return with 8% ROI targets. If the company gives back investors in cash, that cash will be invested at 8% ROI. If the company buys its own stock, its own stock price is expected to return 8% ROI. It doesn’t matter in the end.

We can go into the details to prove this:

Let’s consider a company, with 10M shares, that has a theoretical income of $20M/yr ad infinitum, with its income growing with inflation (We’ll ignore inflation for the time being, let’s just presume that at every point in time they exchange USD for Fun Dollars, where Fun Dollars do not inflate over time, and “$” is the symbol for Fun Dollars). Now, Let’s say this company has liquid capital of $5M on-hand. They have 3 choices:

1) They put the $5M into the S&P500. The $5M will return $400k/yr on average, since the S&P500 returns 8% on average. That brings its theoretical income to $20.4M/yr. At 8% ROI, its valuation will be $255M. That means that each stock will trade for $25.5 dollars, as there are 10M shares outstanding. The next year, it will make $20.4M. That new $20.4M, if invested in the S&P500, will return 8% per year on average, or $1.632M. This brings the income of the company to $22.032M/yr in the following year including the $20.4M/yr it already makes. At 8% ROI, a company that makes $22.032M/yr will be valued at $275.4M. Since there are 10M shares outstanding, the average share price next year will be $27.54 dollars. 27.54/25.5 = 8% Return on Investment exactly, to the dot.

2) They give the $5M back as dividends. That would be a distribution of $0.50 cents per shareholder, as there are 10M outstanding shares. Then, the company will net $20M/yr, giving it a valuation of $250M. This is $25 dollars per share. The way this works is that on the day before the ex-dividend date, when the company still has $5M in cash on-hand, it will be trading for $25.50/share. After the ex-dividend date, the company no longer has $5M in cash on-hand, so the stock will drop to $25.00/share, and the investor will have $0.50 in cash. It’s in essence a liquidation of a partial amount of the stock price. In the end, that means that each investor will end up with $25 dollars in the share, and $0.50 in cash, or $25.50 just like last time. Since the company makes $20M/yr, next year it will have a valuation of $270M: $250M from the income, and $20M from the cash, so the share price one year from now will be $27 on average 1 year from now. That’s not a coincidence, 27/25 = 8%, after the dividend is paid it will have accumulated an 8% return. And, if the $0.50 in cash that the investor got, is invested in the S&P500, it will return 8% on average, so it will be worth $0.54 at the end of the year. In total, the investor will have had $27.54 in the next year, $27 in the company, and $0.54 in cash. 27.54/25.50 = 8%, just like last time.

3) They engage in $5M in stock buybacks. As mentioned in (1), the company if it holds onto the cash itself, its stock price will be $25.50. So, before they buy back the shares, the company will be trading for $25.50/share. Thus, buying $5M in shares will give them 196,078 shares. That means that they will no longer have any cash, but they will indeed still be making $20M/yr in income, and now there will only be 9,803,922 shares outstanding. Next year, the company will have $20M in cash from that income, and still be making $20M/yr. Thus, the company will be valued at $270M next year ($250M from the prospect of continued income, and $20M in cash), and it will have 9,803,922 shares outstanding next year. Thus, each share will be trading for (270 million)/9,803,922 = $27.54 on average. Thus, the investor turned 25.50 into 27.54 on average, so they achieved a - wait for it -, 27.54 / 25.50 = 8% return on investment.

So, no matter what the company does, the investor will still receive an 8% return on investment on average. It literally just does not matter what the company does. At the end of next year, the investor will inevitably have $27.54 at the end of next year on average, and there’s nothing that can be done about that.

So, what happens to the capital in each situation?

1. The money goes into the S&P500 directly as held by the company
   
   
2. The investors get the capital, who will inevitably invest it in the S&P500
   
   
3. On the day that the stock buyback occurs, investors will be left with $5M extra as more net sales were made than net buys, and that money, will inevitably be invested in the S&P500

:%s/S&P500/Your Favorite Hedge Fund/g

So, it literally just does not matter what the company decides to do with the company. It will always make 8% Return on Investment. And the remaining income, will again be reinvested at a target 8% return on investment. And - here's the most important part -, it will always be invested in a company other than itself (To be specific, it'll be in the hands of investors, who will analyze optimal capital allocation and redirect capital to the companies that need it most, of which 99.9% will go to other companies and only 0.1% might go back into itself and only if investors believe that's what should happen). If it chooses (1), (2), or (3), then it admits that it has no profitable way to invest it that beats target ROIs. If it doesn't choose (1), (2), or (3), then it has an investment that it believes is superior, and if investors agree then the stock price will represent that. But (1), (2), and (3), remain identical in all situations.

~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~ Q and A

~~~~~~~~~~~~~~~~~~~~~~~

• (Q) But what about investing in the company? (A) That already happens, the company will do this if management believes that their investment will return greater than 8% return on investment. Investment doesn’t increase the stock price because it’s “built in” to the price, if Uber consistently knows how to turn $10 into $15, then that’s net income of $5 return on investment, which yes it’s 50%, but that income then goes into the valuation of the company which will be 8% return on investment. If management believes that they have cash that will not generate an 8% return on investment, then it will engage in (1), (2), or (3), and it does not matter which they choose.
• (Q) What about companies taking out debt to finance stock buybacks? (A) This makes no sense. Debt will cost the company ~4%/year to take out, and will only expect to return 8%. Doesn’t that mean that the company will profit 4%/year? No, because it has to pay back that debt every year. Investing in the S&P500 involves high levels of variance, you can’t continuously take money out to pay off debt without reducing your principal. If you have ever heard of the “4%” rule, that’s what that is, you can only withdraw 4%/yr out of stock investments and in doing so you prevent the capital from increasing. If you invest $1M in the stock market, but continuously withdraw 4% per year, then your capital over the course of 20 years will remain approximately the same, because your withdrawals prevent gains. See https://www.mymoneyblog.com/wordpress/wp-content/uploads/2017/06/pcharts_me.gif Thus, the shareholders will gain $0 by taking on debt to invest in the stock market, and as mentioned whether the company picks (1) (2) or (3) the capital will always resolve by being invested in the S&P500. Even if the CEO has millions of dollars in shares, he will not see a return on investment by taking out debt just to invest in the S&P500, and as a shareholder it doesn’t matter (1) (2) or (3) the capital will inevitably be invested in a stock that returns on average 8% per year. The shareholders do not want this, the CEO does not want this, the CFO does not want this, so this is why this does not happen. Just because a company has cash, debt, and buybacks, doesn’t mean that the debt went to the buybacks. Instead, the cash went to the buybacks, and the debt went to the investments of the company. This is only situation that remains profitable for shareholders, the CEO, the CFO, et cetera. If the company can withdraw debt at lower rates and profit, should it? Yes, absolutely. This will happen regardless of whether the company chooses dividends or buybacks. Companies like AAPL which had $100B+ in offshore accounts could trivially take near-0% interest loans and invest them, while still paying out dividends the entire time, and capital intensive industries such as DAL can collateralize their assets into liquid cash. Too much debt lowers the acceptable p/e ratio, so of course it's all done in moderation, and the best loans have collateral so it's more of an "exchange" if you will. And they will do this regardless of whether or not it's a dividend or a buyback.

~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~ So why all the complexity, why not just use dividends

~~~~~~~~~~~~~~~~~~~~~~~

So, why do companies use buybacks instead of dividends? Let’s look at all three options actually:

1. Using (1), the money is forcefully invested in the S&P500. This is irrational and restrictive, because while the S&P500 might return 8%, it only returns 8% because of the decision making of investors that influence stock prices into equilibrium, the system only works when the money is back in the hands of investors who will eye down the next target investment. It’s in general simply more rational to give the cash back to shareholders using (2) or (3) so that they can engage in re-investment. If they wanted to buy SPY, they can buy SPY themselves.
2. In (2), the dividend will be distributed to shareholders. This is a little frustrating from the accounting perspective, because of various tax analysis with dividends, and because reinvesting dividends results in discontinuities if the dividends can’t buy back a whole number of shares back into the S&P500, so the investor frustratingly needs to participate in senseless accounting practices such as fractional shares and fractional dividends to experience their full return on investment and continuously compounding effects. Additionally, it creates discontinuity in stock prices that are needlessly complicated. Throughout the quarter, the stock price of a company that gives out dividends will slowly rise from $25 to $25.50 over the course of 3 months. Then, at the end of the 3 months, it will abruptly drop from $25.50 to $25, and the investors will get $0.50 in cash that will be reinvested in the same stock or the S&P500 regardless. This creates needless complications when analyzing option and futures markets, complications with short selling and having to pay dividends back to the owner you borrow your share from, and having to adjust volatility indexes to account for the discontinuity.
3. With (3), Stock buybacks create a simple case for investors. They pay $25.50 for a stock that will be worth on average $27.54 a year from now, an 8% growth in the stock price.

Everything that actually reflects reality between (1), (2), and (3), will be in general identical. Between (1) (2) and (3), you still get an 8% return on investment, it’s just clear that from an accounting perspective (3) is just the easiest and least intellectually demanding to understand, while still freeing the capital back into investors hands unlike (1). It avoids having to implement more ex-dividend algorithms and after-hours algorithms to handle the ex-dividend date, and avoids having more traders dedicate their time to ex-dividend analysis, because it just isn’t an efficient use of time. With (3) it’s simple - every stock will gain 8% on average. And sure, the company doesn't care about the traders spending extra time, it's still in general easier for the company to just buy the shares outright with a stock broker when it gets the same thing done.

For a clear-cut example of this accounting advantage, I ask a question: If you had $100k in 1990, and invested in the S&P500, how much would you have today? You might think you should take SPY in 2020, SPY in 1990, and then divide the two, and then multiply by $100k. But that’s wrong, because SPY returns a 1.94% dividend. So now the mathematics is annoyingly more complicated, and you need to extract dividend history for each quarter of SPY between 1990 and 2010 and integrate that into your calculation. It requires spreadsheets, it requires more python programming, it requires dataset extraction from APIs for dividend history. It’s beyond ideal, there are calculators online, but what about for every other stock, what about for algorithms on quantconnect.com that have to integrate these calculations? If all companies engaged in (3), to calculate your return, you just take the stock price in 2020, and divide by the stock price in 1990. It’s unbelievably simple. That’s exactly what your return would be. You can look at a stock's price history to see exactly what $1 invested at time X would have netted you today.

It's just simpler, No more need to build-in dividends into option prices and stock prices for the discontinuity. When paying CEOs in options and stock, the valuation of these options and stock are easier to calculate, it's all easier. There’s just less mental effort needed. That’s all. That’s basically it.

~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~ How much of this is simplified?

~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~~~~~~~ ~~~~~~~~~~~~
Side-Rant 2

• This is why it’s very horrifying when bureaucrats want to get involved in capital allocation. Rather than money rewarding people who are correct and punishing people who are wrong, bureaucrats can wave their magic wand and declare something to be true or false if enough people believe in them. So much rhetoric involved in “But they should reinvest instead of buybacks!” Should they? If they can beat 8% then yes, if they can’t then they should give it back to investors and let the investors find companies that can beat 8%. Rhetoric like “have the companies reinvest” will “feel” good in that you might “feel like” companies will invest more, when in reality you’re screwing with a consensus algorithm that will naturally guarantee efficient allocation of capital regardless of who believes what. You’d force Delta to invest into …. What? More planes? They already have planes. Pay off debt? What if their capital-secured loans have interest rates lower than expected returns? The fact that Delta has planes means that they can keep them as collateralized loans that are low risk as creditors can seize the plane if Delta fails to pay, and this doesn’t affect the economy because the creditors will simply sell that plane to another airliner who will keep that plane making the absolute maximal amount of money possible - and thus contribute to society in the absolute maximal possible amount (With competition cutting margins down to the cost and capital of doing business itself). Stock prices care about raw expected value, not risk of bankruptcy, if things go sour they reorganize under Chapter 11 and go back into business keeping what’s profitable, yes investors lose all they money but that was the risk when eyeing down optimal levels of returns, and the investors get to move capital out of Delta and to where it’s needed such as companies like AMZN/NVDA/AMD that will continue to grow, they will do anything in the whole world that can optimize investor returns on average, anything. The government uses rhetoric that implies that they want to shove capital down Delta’s throat by blocking buybacks, while suffocating innovation such as Uber and Airbnb from the capital that they need, as Delta tries to figure out what to do with its money because the government tried to tell it that the government knows best, mostly because the word “buyback” is emotionally unfamiliar meanwhile the word “dividend” sounds much more pleasant. The idea is built entirely on emotion, when in reality the market will naturally punish people who use emotion to trade stocks, so only the beautocrat has the opportunity to turn emotion into reality. The emotional trader will simply blow up his account, and never become anything more than a blip. In reality, at the end of the day, most people find the word “dividend” emotionally acceptable, so even if buybacks were banned, it’s not like Delta would invest, they would just pay a dividend, since that’s still optimal capital allocation. “Dividend” “feels” like a return on investment, so bureaucrats would not longer be able to claim “They spend $13B on buybacks!” anymore, and Wall Street just needs to do a little bit more paperwork and dedicate a slight bit more mental capital towards working around dividends again. Not a big deal, but it shows that in the end the government will never win the “Do this, do that, you should do this” debate, the micromanagement of companies by the government will never optimize productivity or even move it in any positive direction. Other than certain policies like taxing income, which disincentivizes investors from actually ever withdrawing their investment, which is debatably a positive outcome. It's all game, it's all incentive structures, as long as you build a system that gives the most money to the best capital allocators, don't tell people what to invest it, the system will handle itself, if you want to move things in positive directions all you have to do is build the correct incentive structures that do what you want to get done, not micromanage, micromanaging is never the solution. No, you do not not know best, don't say you know how to best allocate capital. Not unless you're an investor yourself. In fact, nothing makes an investor happier than improper capital allocation, since then he himself will reap the rewards of the proper capital allocation, correcting the market with his own actions. It's an interesting system, either (a) it has the properties that you want it to have such as "all stocks return 8% on average", or (b) it doesn't, in which case you can profitable massively off trading based upon your analysis. A beautiful system, the harmony of the trades and the bell that rings with every close. 

~~~~~~~~~~~~ ~~~~~~~~~~~~