Hacker News Reader

Comment by cperciva | Hacker News Reader

Remember, in a perfectly liquid market without taxes, a company which executes a share buyback will have no change in its stock price, since the shared redeemed will be exactly balanced by a reduction in the value of the company; and this is also exactly equivalent to distributing profits in the form of a dividend.

On the other hand, when the rate of taxation on capital gains is lower than the rate of taxation on corporate dividends, it's advantageous to pay out surplus cash in the form of capital gains by executing a share buyback. This also provides a tax-planning benefit to shareholders: Everybody is forced to realize a dividend (and pay tax on it) but some shareholders will prefer to realize more or less capital gains in a given year.

In Canada we have rules which sometimes reclassify capital gains as "deemed dividends" to prevent this sort of maneuver. A far better solution would be to simply fix the tax system so that economically equivalent actions get taxed identically in the first place.

View on HN

> a company which executes a share buyback will have no change in its stock price... this is also exactly equivalent to distributing profits in the form of a dividend

This is false (conditionally). Most valuations look something like: earnings x growth + cash. If you have cash that you cannot reinvest at the same ROI that you have been growing at, you can increase shareholder value by removing the cash element, since it's dead weight. A buy back is a way to invest that cash at the same ROI that your company is growing at. A dividend would require the investors to re-invest that cash and get the same ROI that the company is growing at.

This is why it make sense for large, fast growing companies to buy back their shares and for large, slow-growing companies to issue dividends.

Finance theory disagrees with you.

A dividend lets the investors choose if they want to buy more shares (from the same sellers who would sell in a buyback) or mix their current shares with the cash.

Investors buying more shares with dividends doesn't decrease the number of shares outstanding. If you increase the share price without decreasing the share count, you have created value out of thin air and are saying the company was more valuable without the cash than it was with it.

The key is the asymmetry of information. The market price should reflect all public information. However the company has non-public information, which puts them in the best position to judge if money is best invested inside the company or out. Share buybacks are a legal form of insider-trading.

Number of shares doesn't matter, enterprise value does and it doesn't change. Your implying that there would be a further equity issuance which doesn't routinely happen on buybacks.

Exactly. If number of shares would matter, share splits would matter.

Market cap = shares outstanding * share price.

Reducing shares increases the price (while keeping market cap the same). Share splits increases the shares outstanding and decreases the price. So... yes, shares outstanding does matter.

Now you are confusing outstanding shares with market float. Shares repurchased become treasury stock and are now apart of equity.

Whether the company destroys the shares or retains them as treasury stock is irrelevant. Investors don't care if their share price went up by 10%, or if they effectively have 1.10 shares.

> Investors buying more shares with dividends doesn't decrease the number of shares outstanding. If you increase the share price without decreasing the share count, you have created value out of thin air and are saying the company was more valuable without the cash than it was with it.

You're wrong. When the company issues the dividend, the share price falls for precisely the amount issued. So if everybody used that money to buy the shares again, the share price should return (roughly) to the value before the dividend was paid. The number of shares outstanding wouldn't change.

You're wrong. The market cap already had the cash priced into it, regardless of how it was used (as a dividend or re-invested). By your logic, a company whose profits remained flat and issued a dividend would become less valuable every year until it was worthless.

No, you are missing the point.

There is a value in a future dividend. On the ex-dividend date, the share becomes less valuable because the new holder will not receive a dividend in X days. This X is usually very small, so a share that goes from paying you $5 in two weeks to not paying you has clearly lost value close to $5.

The market cap already reflects the value of all future earnings discounted to the present. The timing of dividend payments is irrelevant.

If your argument was true, that a stock's price predictably fell the day after the dividend, investors could simply short the stock and get a guaranteed profit, which is not possible in an efficient market.

> The timing of dividend payments is irrelevant.

False. If I receive cash today I can reinvest it and start earning a return. If I receive cash in a month I have forgone one months reinvestment return.

> investors could simply short the stock and get a guaranteed profit, which is not possible in an efficient market.

False. Well if you are short a stock over ex dividend date then you need to pay the owner of the stock (whomever you borrowed from) 1) the dividend which he has forgone 2) a financing spread equal to a benchmark (e.g. FED Funds + 300 bp's) for the duration you are short.

No offense but you really haven't thought this through very hard.

Also markets are not efficient despite what you read in academia.

The net present value takes into account the difference between payment today and payment tomorrow.

The ex-dividend rate is an implementation flaw that makes the stock price discontinuous at the dividend date. If dividends were pro-rata it would not be.

I never said markets were perfectly efficient.

You didn't think through your answer very hard did you? ;)

I think we are going to continue disagreeing. You seem to enjoy Finance so I would encourage you to learn more about it for your own benefit.

> that a stock's price predictably fell the day after the dividend,

They do. But you don't usually see this, because the prices are "adjusted", to make price history continuous, same as with stock splits.

Take a look at MSFT (Microsoft) stock price on Yahoo! Finance. "Close" is what actually happened, and "Adj Close" is what you see on the graph.

  Date	          close      Adj Close*
  16 Feb 2016     51.09	     51.09
  16 Feb 2016                             0.36 Dividend
  12 Feb 2016     50.50      50.14

  * Close price adjusted for dividends and splits.

https://uk.finance.yahoo.com/q/hp?s=MSFT

> investors could simply short the stock and get a guaranteed profit

No, to short a stock you need to borrow it first. If you borrow a stock, you owe the dividends that were due in the meantime.

The adjusted price reference you made is axiomatic. Adjusted prices simply subtract the dividend. It does not reflect market value. You'll notice the open on the day after the dividend, the stock actually went up $0.40 from the closing price.

When shorting, you only pass the dividends through. The company sends you the dividend, you send it to the original owner, you could still profit from the drop in share price.

Well, the stock price might have changed because of some news or general market sentiment. I guess I could have found a better example, a stock where the dividend is much bigger (percentage-wise), the effect would be much clearer then.

Edit: e.g. EWM (an ETF, I think)

https://uk.finance.yahoo.com/q/hp?s=EWM&d=3&e=4&f=2016&g=d&a...

No, shorting doesn't work that way, obviously. "Shorting" means you sell he stock, so the new owner gets the dividend, not you. That's one reason shorting equities is very risky long-term.

The ex-dividend date is an implementation flaw that makes the stock price discontinuous at the dividend date (I initially misunderstood this). However, in theory, a pro-rata dividend would be continuous.

Yes. That's what happens with bonds (clean vs dirty price). It's more complicated with equities because the size of the payout isn't known in advance.

No, it happens just before the dividends.

Say you own 1 share. It's trading for $20. Tomorrow the company pays $2 dividend. Then your share is worth $18 and you have $2 cash as well.

Alternatively, if you're buying the share, you're willing to pay $20 for it today but only $18 tomorrow, because you know that you won't be getting a $2 dividend if you buy it.

If you take future dividend payments into account, you also need to discount them. If you think that (discounted future dividends) > (stock price), that's a signal for you to buy. If enough investors reason this way, the price will rise until (discounted future dividends) ~~ (stock price).

The market cap already reflects the value of all future earnings discounted to the present. The timing of dividend payments is irrelevant.

The future earnings have huge uncertainty to them. A company with a $50 stock price might have $5 per share worth of cash in the bank, (which is something you can actually observe.) The other $45 represents expected future earnings. If the stock pays out a 50c dividend, the stock now has $4.50 per share in the bank and the $45 expected future value is unchanged.

Either you're stupid, or there's a misunderstanding. Probably the latter. Let me try again. Assume that δ is the discount factor (δ = 1 / (1 + r)), P is price, and D_i is the dividend in year i.

Before dividend:

  P = D_0 + δ D_1 + δ^2 D_2 + δ^3 D_3 + ...

After dividend:

  P = δ D_1 + δ^2 D_2 + δ^3 D_3 + ...

When calculating the value of an annuity, all that matters is your number of periods, the amount, and your discount rate. This is why an annuity appreciation chart over time is continuous.

This comment would be better without the gratuitous namecalling.

Sorry, I get very (too) emotional when someone is wrong with math...

Profits being flat isn't quite right. Profits being zero is the case where a company issuing a dividend would eventually become worthless.

This is a strawman argument. Cash is not required to be re-invested to maintain current profitability.

That's not the argument I'm making. I'm not disagreeing with you.

The point I'm making, which I admit is an obvious one, is that without profits a dividend will erode a company's value. Simple flow of money. Unless there's a surplus flow of money in, you can't have an outward flow. Therefore, the effect of a dividend on a company's market cap depends on how much profit they have in comparison.

Or at least if management thinks (or wants the investors to think) they can continue to exceed the ROI that a normal investor would get making their own decisions.

I don't think this is accurate. When a company buys back shares, the shares are retired increasing the ownership percentage of the remaining shareholders.

It is basically the opposite of issuing shares and diluting existing shareholders.

Right. The remaining shareholders own a larger percentage of a company that now owns less cash.

Say the only thing my company owns is a bank account with $100 in it. There are 5 shares outstanding worth $20 each. The company buys back one share for $20, so now there are 4 shares outstanding in a company that owns $80.

People don't invest in companies that merely have cash. Growth and recurring revenue are more important metrics to most

OK, but if the company held $50 in non-cash assets and $50 in cash, the same logic would hold. The "growth and recurring revenue," stuff you're talking about is just an asset pricing restriction, which requires that the $50 in non-cash assets are not affected by moving some cash off the balance sheet. This holds if the company does not face cash constraints. (Like Apple.)

It depends on the industry. Some stock valuations show that investors in certain industries like companies to hold on to cash reserves. This is especially true in the technology space where investors want companies to be able to cash in on the next big thing. Utilities, though, see their stock punished for holding onto excess cash.

Well in theory, if your shares are diluted /2 but the company's capital is x2, and all that capital is used efficiently, you have not lost anything either. Half slice of twice the pie.

That makes sense. It is true the shareholders will own a bigger piece of a smaller pie.

Both situations are equity+cash=equity+cash.

The ownership percentage is hidden inside the equity.

What's not accurate?

There's another issue created when companies issue debt in order to fund share repurchases. It may seem to make financial sense rates are so low but it's created a huge amount of additional leverage throughout the corporate world.

That's because interest payments on debt are taxed at a lower rate than dividends. Fix the tax system, and the leverage will go away.

This is not true. You are ignoring the most important factor of all, which is whether the company's share price is above or below its "intrinsic value". Of course, intrinsic value is a fundamentally fuzzy concept, but in some cases the share price is obviously out of line relative to historical measures of valuation (price to tangible book value, TEV/EBITDA, price to sales ratio, etc.). If there is no good fundamental reason for the lowered valuation (loss of patent protection, entrance of new competition, secular changes in technology, etc.), then it could create value for remaining shareholders if the number of shares is reduced and that capital is replaced with debt (or simply spending excess cash on the balance sheet). If the share price is above a conservative estimate of intrinsic value, then a company can easily destroy shareholder value by buying back shares. In that case, they might be better off issuing more shares or doing an acquisition for stock to take advantage of their overvalued currency. This sort of capital management is one of the most important ways a management team and board of directors can create value over time. For a good historical example, look at the case of Teledyne ( http://www.capitalideasonline.com/articles/index.php?id=2725 )

Intrinsic value does not exist. Which is why we have markets in the first place.

The obvious way to demonstrate this is you sell 10 shares for a higher price than 10 billion shares.

It's true that intrinsic value is not usually well defined (except in simple cases, where a company is going to be liquidated and has clearly defined assets and liabilities that can be precisely measures, such as cash/marketable securities), but that doesn't mean that a thoughtful investor can't determine a reasonable range of likely values that errs on the side of conservatism. For example, by zero-valuing assets that are hard to quantify, such as an undeveloped plot of land. Or if there is an annuity-like cash flow stream from a long-term contract, one can discount those cash flows to the present using a conservative (high) discount rate. In cases where the share price is way below a conservative downside-case valuation, it's not necessary to know with precision what the intrinsic value might be.

Individual investors can define a proxy for intrinsic value. But, overall investors will come up with different numbers. Thus investor A may think a buyback was a net gain, while investor B may think a buyback was a net loss.

The calculation is (net income - preferred dividends/ weighted average common stock outstanding) = eps. Eps is on the income statement, face of financial statements and will go up if wacs goes down. Common stock out standing doesnt include treasury stock, stock repurchased by the company.

But none of this explains the main reason they would do this (a buyback or a dividend) instead of reinvest the profits. Yes the have lots of cash, but they have for awhile.

So the interesting question is why are they doing this now given the medium/long term ROI a company with cash looks generate. It could mean they stopped seeing obvious medium/long term investments. Maybe a small bubble is 5-10 years away.

To see the answer to your specific question you need to look and compare these graphs:

1) https://ycharts.com/indicators/sp_500_eps (note: slowly rising, if you ignore seasonality)

2) https://ycharts.com/indicators/reports/sp_500_earnings (note: dropping fast)

TLDR earnings are going down, but earnings/shares are going up. So what is going on ? Earnings for the US economy as a whole are dropping (pretty fast even). But the metric investors use to value shares, earnings per share is going up.

That means U.S. companies are buying back shares at a faster rate than their earnings are dropping. Why ? Exactly to generate this outcome : normal valuation metrics for shares (net-present-value of future earnings per share) go up as a result of this operation. When cutting a million corners the share price of any (large cap) stock should be roughly NPV(8%, future_cashflow).

The next question to ask is ... given that this uses a LOT of debt that is currently at very low interest rates, what happens if interest payments inevitably go up (either as a result of inflation, or of the FED raising rates) ? The problem with low interest rates is that, at the moment, 1% rate rise would quadruple interest payments for the government, and double interest payments for AA corporations.

I might be wrong here but aren't most US companies flush with cash these days? I would assume that they use their cash reserves for the buybacks and not debt. Are there any stats on what is used to finance the buybacks?

Most US companies have cash overseas and have never bought them back to the US because doing so will incur a third gone in taxes. Thus they are lending the cash to the US entity to execute but backs, hoping that a tax holiday is declared in the future.

> they are lending the cash to the US entity to execute but backs

I'm pretty sure they don't do this as a loan from the untaxed or under-taxed overseas entity back to the taxed parent would create a tax event.

AAPL has more cash than it knows what to do with, but it's a huge borrower in the corporate market so that it can fund its dividend payments.

It's trivially true that not all companies can provide ROI surpassing the market average. The more interesting question to me is why they ever stopped.

Even if the tax percent was exactly the same on capitals gains and dividends, buybacks would still make sense. Dividends typically are paid out at least once year. On the other hand, you only incur capital gains when you sell shares. So, with share buybacks, you can potentially wait many years before you pay the tax, having the money accumulate for you in the meantime.

Don't forget the signaling effect, which can significantly boost share price

Dividends also signal.

and this is also exactly equivalent to distributing profits in the form of a dividend.

I don't see this equivalence. I guess It's rather like settling debt since after a buy-back there will be fewer future dividends to pay. But nothing changes for continuing shareholders, does it?

Assume an investor takes all dividends and buys more shares.

Example:

Assume 10M shares at $1 each. With a 10 cent dividend ($1M total), an investor with 100k shares (1% of the company) receives $10k. The market cap drops to $9M (since cash holdings decreased) and the stock price drops to $0.90. They then buy ~11k (10k/.9) shares and have 111k shares, or 1.1%.

Now instead assume the company does a buyback of 1M shares. Now there are 9M shares, each still priced at $1 (cash holdings decreased). The investor has 1.11% of the company.

Lots of assumptions here (no fluctuation in price due to market reaction, cash is not discounted in market cap, etc) but the numbers check out in a perfect world.

The point is that with a Dividend investors can choose whether to reinvest, whereas with a buyback everyone essentially reinvests.

Well but that just confirms my whole point. In a buyback the company invests its cash into stock instead of distributing it as dividend. It therefore makes an investment decision for the investor, who never gets to see any cash.

The whole point of giving out dividends as a company is saying to an investor "here, you can invest this cash better than we can". If you're not giving out dividends you think you have better investment foresight than your investors, which may often be true.

What you as a company then do then with the cash, whether you're investing it in R&D, personnel, etc.. or buy stock doesn't change the mathematics of things for the investor, even if the stock you buy happens to be your own stock.

Yep. Except for the fact that without taxes the investor can just sell the stock if they disagree with the investments the company is making, resulting in the same cash return as a dividend.

In my example with the buyback, if the investor sold 0.11% of the company after the buyback, the result would look just like what would happen if they did not reinvest the dividend. 1% of the company and $10k cash.

Ok, but it's not the buy-back that increases the per-share value for the selling investor, it's the fact that the company has generated positive Free Cash Flow that it is free to invest.

True, though investors often discount cash on hand ("a bird in the hand is worth two in the bush") because of the uncertainty of ever seeing that money. YHOO is probably the easiest example as of late, with their massive goodwill write downs of acquisitions. So the act of either a dividend or a buyback removes that uncertainty and can result in an increase in per share value.

For simplicity, assume that the company buys back stock from all shareholders equally, and then the shareholders trade amongst themselves.

Dividends and buybacks are simply a way to get money from the company to the shareholder. The total number of outstanding shares is of no import---as you can see in stock splits.

Also, companies that borrow money to finance stock purchases (eg Apple) also save on their taxes because interest expense on debt is tax deductible.

Expensing interest only saves you the total interest x interest rate. So they just get a 35% discount on their interest rate.

You're absolutely right that the savings on interest is post tax. However the tax deductibility of interest creates a substantial tax shield and in general, has a net positive value for the company since changing the capital structure by increasing debt decreases the weighted average cost of capital. Studies suggest that debt, because of the deductibility of interest, on average leads to ~1 dollar of additional company value for every 10 dollars of debt financing (vs equity financing).

Total interest * [marginal] tax rate, not times interest rate.

(I give you credit for knowing the difference, but not every reader will know the difference, so I'm trying to help them, not nitpick you.)

I like this idea and my understanding is that they have a similar rule here in Germany.

Reply on news.ycombinator.com