« What is a "cashless" economy anyway? | Main | Money demand and supply in a red/green world »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Fascinating post.

"But the capital gains from share buybacks (or capital losses from dilution) are an alternative to paying dividends (negative dividends)..."

I'm parroting what financial economists might say, but is it the case that share buybacks create capital gains? If in the process of buying back shares the fund manager causes the fund's shares to rise above fundamental value, then deep pocketed shareholders will undertake the opposite transaction, thus negating the capital gain.

Do I have this right?

The Swiss National Bank issues a relatively "fixed" amount of money. This is similar to a closed-end mutual fund which begins with a fixed amount of money and then buys assets. In Switzerland, the (relatively) fixed amount of money can purchase any Swiss asset which is for sale. There is a similarity to closed-end mutual funds here.

Next, Swiss Commercial Banks operate an open-end mutual fund. Because they can only use Swiss National Bank money, the commercial banks trade Swiss National Bank money for bonds. No new money is created here; we only have an exchange of Swiss National Bank money for bonds.

Now let's model that I am a North American wheat grower. I really like the idea of having my money in the form of Swiss Francs. How can I manage to exchange wheat for a Swiss bank account?

I can sell wheat to Switzerland in exchange for a Swiss bank deposit. By doing this, I undercut any price that a Swiss wheat grower is willing to accept (imported deflation). This transaction does not change the amount of money that the Swiss National Bank has in circulation.

We could extend this example to obtain a universe of next actions of the Swiss wheat growers, the North American deposit holder, and the banks involved.

JP: I think you misunderstood Nick's language.
Company A has net assets worth $1 bn, and 100 mm shares out. Company A is worth $10/sh.
Company A gets a windfall of $100 mm. This takes net assets/share up to $11. It could pay a dividend of $1/sh and be worth $10 again on ex-div day. Or it could buy back $100 mm of stock at fair value of $11 per share, and be worth $11 ($1 bn / (100 mm shares -$100 mm/$11).
Once you've decided to share that profit with investors instead of retaining it on the balance sheet, the choice is between giving them dividend income or a higher stock price. It is true that your shareholders get the capital gain even if they don't buy back the shares, however, as long as there's no dividend.

JP: Thanks!

Suppose the mutual fund announces it has earned unexpected profits on its assets, and announces it will spend those profits on share buybacks rather than increasing management expenses. The fundamental value of the shares (and the market value) will rise (by the same amount, assuming efficient markets). But you would get the same rise in fundamental value (maybe not exactly the same, it depends) if it announced it would reinvest those profits instead.

I'm parroting what financial economists might say, but is it the case that share buybacks create capital gains?

There is some theory involved as well, it's not just a saying. Suppose there's a corporation that is in the business of owning gold bars. At today's market price the stash is worth $1 million, and there happens to also be 1 million shareholders, so we could value each share at $1 on the basis that this much gold is owned by each shareholder.

Company directors decide to sell half the gold which gets them $500k in cash, and they buyback 500k worth of the $1 shares. Now you have 500k remaining shareholders and the market value of the gold stash is exactly $500k so the shares are valued at $1.


OK, different corporation owns a productive coal mine, valued at $1 million, also has 1 million shareholders. The coal mine turns a profit and makes $150k in a year after operating expenses, but the management decides that $50k needs to be spent on capital replacement because some equipment is wearing out. That leaves them with $100k in cash they have no particular use for, so they could [A] pay a dividend to shareholders of 10c per share, or [B] start a buyback program.

For case [A] every year the share price will creep up until it is $1.10 just before the dividend pays out, then it will drop to $1.00 again right after the dividend pays, and creep up for next year's dividend. This is logical, because the longer you own that share, the more of the profit you would be entitled to. To reduce the saw-tooth some companies pay a bundle of small dividends spread over the year, which has the same effect.

For case [B] the operation of the company is exactly the same, the return on investment is exactly the same, it's just paid out via a different pathway. However, in case [B] the total number of shares is reducing while value of the coal mine stays about the same, so existing shares *SHOULD* increase in value by exactly the rate of return of that productive investment. Once more we get to the point where the longer you own the share, the more of the profit you would be entitled to. Let's suppose we earn a consistent 10% every year, so after 7 years and 3 months the company is down to 500k shares, each worth $2 and management decides to split the stock... back to 1 million shares worth $1 and right where we started from... no different to paying dividends.

There's no reason for a share back to affect the price of the stock, other things equal. It is an exchange of cash for stock. The post-buy back stock has a higher expected return but is riskier, both effects due to the removal of cash from the balance sheet. The only valid reason to do a share buy-back is to give remaining shareholders a riskier stock with a greater expected return - in the absence of equivalent risk new investment opportunities.

I must say, I don't follow 100%, but this seems not quite right:

And, unless the Swiss impose exchange controls (which they don't want to do), this means the Swiss National Bank must satisfy that demand for Swiss Francs, because if it fails to issue enough Swiss Francs to satisfy that demand, the fact that the Swiss Franc is used as medium of account and medium of exchange within Switzerland will screw up the Swiss economy.

The former chairman, Philip Hildebrand, tried out the 'Chuck Norris peg' of the CHF to the Euro. It worked for a while but was finally brought down mainly due to political pressure from the 'End the Fed' right wing who argued it was taking on too many risks and shouldn't actually be doing anything (your favourite theme, I gather). That political pressure then brought on the speculators and led to an expansion of the SNB balance sheet. So their prophecy was self fulfilling.

Under his successor, Thomas Jordan, the peg was lifted. And since then the CHF has appreciated substantially against the Euro. That DID srew up the Swiss economy and Switzerland has reported higher unemployment rate than Germany for the first time since I can remember. Probably since WWI. Exports, especially in machinery, are down a lot.

The exchange rate has come down a bit recently, maybe due to less overt FC operations by the SNB. Seems under Jordan the SNB has refocused on traditional Swiss values of shutting up and making money.

Another thing is that the SNB actually pays out dividends to the Cantons once a year - if it makes a profit on its investments, that is. And it is not limitied in the kinds of assets it can buy. But, unlike a mutual fund, its mandate does not include aiming for a profit. As far as I know, that is quite a unique setup for a central banks.

The idea of a sovereign wealth fund, like that in Norway, has also been floated.


"But the capital gains from share buybacks (or capital losses from dilution) are an alternative to paying dividends (negative dividends), so that problem with the metaphor shouldn't matter much."

What does Modigliani–Miller tell us about a dividend policy?

JKH: suppose you do a share buyback instead of giving the cash to the government (or giving the managers a Xmas present). That will affect the share price. And that is what we are talking about here. (Though there is also a separate effect if the demand curve for shares is not perfectly elastic wrt the rate of return, and the demand for money is not always perfectly elastic wrt rate of return, witness people still holding some currency despite extremely negative rates of return in hyperinflations.)

Oliver: simplest way to think of it: If the SNB does not satisfy demand for CHF, the exchange rate appreciates, and demand for Swiss net exports falls, so Switzerland hits recession.

Avon: Is this an exam? MMT says it doesn't matter at all.

So the key isn't buybacks, it's how to divvy up profits between shareholders and management. A decision to vary the management expense ratio will cause a change in the unit price. But all things staying the same, a decision to keep shareholder cash on the balance sheet or spend it on buybacks is irrelevant with regards to the unit price.

Vice versa, management can't dilute shares by printing loads more shares, insofar as the assets that it buys are being purchased at their fundamental value and the MER stays constant. IOW, open market operations are irrelevant. Only the MER is important.

Nick, that I know. I thought you were saluting the SNB for not having let the CHF appreciate. My point was, it has. My misunderstanding.

The share value does not rise with a buyback.

It drops with any added expense, MERS included, other things equal.

A buyback is not an expense.

JP: "So the key isn't buybacks, it's how to divvy up profits between shareholders and management."

Yes, (though by "management" we really mean "the government", which owns the central bank. And what is the right word for the owners of the mutual fund itself, as opposed to those who own the units??)

"But all things staying the same, a decision to keep shareholder cash on the balance sheet or spend it on buybacks is irrelevant with regards to the unit price."

Well, that is nearly true, but it won't be exactly true. Suppose this mutual fund is more liquid than other assets. The marginal liquidity premium may be a function of the number of mutual fund shares. The money demand curve is not perfectly elastic at any given rate of return. It slopes down.

Ok, got it. I'm on board.

I see these kinds of metaphors silly. Portfolios of assets have an expected return based on the premium for holding risk. The decomposition into the factors to which the portfolio is exposed will determine how high that expected return is. That observation has little to with this post. In fact, it's down right silly.

I'm afraid I have to agree with Avon Barksdale in this case.

Metaphorical explanation is a curious thing.

"You can, if you want, think of (central bank) money as shares in a closed-end mutual fund."

This has been your current post for the last 3 days and I have continued to think about it.

My thoughts have brought me to think that the assets accessible by the mutual fund shares (in this metaphor) are the entire assets of the nation. Accessibility is not the same as ownership because no one would say that ownership of money constitutes ownership of the nation's assets.

It would be correct to say that anyone who owns money (shares in the mutual fund) does have the ability to trade those shares for any other asset including financial assets. Both money and assets can be freely traded.

How can the number of shares in a mutual fund be increased? Only the owner of a mutual fund would have the ability to create new shares in that fund. The central bank has the ability to create money so we could say that it is creating new, additional, shares. These new shares have the same rights as older shares.

Next we could examine how the new closed-end mutual fund shares flow into the hands of prospective owners. I think this examination takes back into more familiar models of money supply expansion and the expected sequence of successive beneficiaries.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad