Scott Sumner  

Apple's open market purchases

PRINT
Murphy on Interpersonal Utilit... The Portuguese Puzzle: Decrimi...

All transactions are two-sided. However in most cases the market for the object being sold is impacted much more strongly than the market for the object being purchased. If a huge new silver mine started selling silver on the open market, the impact on silver prices would be dramatically larger than the impact on the market for whatever they receive in exchange, unless they engaged in some sort of barter. The same is true of sales of Boeing airliners. Boeing's sales have a big impact on the airplane market, but not much impact on the market for whatever asset is used to buy their jets (say money or bonds.)

From January 2002 to January 2006, the US had a big housing boom (some call it a bubble, I don't.) During this period the Fed steadily increased the money supply, boosting the monetary base by over $9 billion per quarter, a total increase of $149 billion over 16 quarters. And the economy was smaller then, so that would be equivalent to an increase of $12 to $14 billion per quarter today. These injections (OMPs) were used to buy securities---mostly Treasury securities.

To really understand monetary policy you need to think about what makes the Fed unique. Here's another entity that is buying lots of bonds:

The tech giant reported record profits last quarter of $18 billion - the most of any S&P 500 company ever. Apple's blowout quarter single-handedly pushed overall S&P 500 earnings growth into positive territory after the company reported Tuesday night. As of then, without Apple, earnings were projected to contract 0.5%, according to FactSet. . . .

Also standing at record levels: Apple's cash and marketable securities. That figure ballooned to $178 billion last quarter, up from $155 billion in the prior period.


So Apple is probably buying up even more securities than the Fed did during the housing boom, even as a share of GDP. And of course Apple is hardly unique; lots of other companies buy bonds, and countries such as China and Japan have accumulated far more securities than even Apple.

Because the global debt market is enormous, and because many types of debt instruments are close substitutes, Apple's purchases are not very important. A giant silver mine, or Boeing, or Apple, is important because of what they sell, not what they buy. They have enough market power to influence the price of what they sell, but not enough to significantly influence the value of most securities. (Obviously this would not hold if they accumulated just a few thinly traded stocks.)

And the same is true of the Fed. But wait (you must be thinking) don't Fed OMPs of Treasury debt cause debt prices to rise? Not always, massive debt purchases in the 1960s and 1970s caused bond prices to fall sharply. But yes, in many cases it is true that Fed OMPs do cause bond prices to rise. However that's not mainly because of what the Fed is buying, it's because of what they are selling. Even when central banks used to buy gold instead of Treasury bonds, interest rates tended to fall immediately after monetary injections, due to the liquidity effect. Until prices adjust, interest rates must fall until the public is willing to hold the larger cash balances. So the direct interest rate effect of Fed bond purchases is trivial, just as the interest rate effect of Apple's bond purchases is trivial. It's the money injections that matter, and interest rates may rise or fall on those injections, depending on whether the liquidity or Fisher effects dominate.

Does this change at the zero bound? Perhaps, but not necessarily in the way that you'd assume. Over the past 6 years the Fed bought far more debt than did the ECB, and that's why longer-term risk free interest rates are higher in in the US than the eurozone. The Fed policy led to faster NGDP growth in the US than the eurozone.

Central banks are unique because of what they sell---they have a monopoly on base money production---not because of what they buy. There's a reason it's called "monetary policy" and not credit policy.

PS. It's possible that the huge Potosi silver mine depressed the value of money in the 1500s, but that's because in those days silver was money. It's the supply and demand for the medium of account that matters; what they buy with the new money is of trivial importance (unless it's something with a much more thinly traded market than bonds.)

PPS. When the financial press says "cash and securities" I believe the term 'cash' generally refers to short-term debt such as T-bills, not actual currency. Someone please correct me if I am wrong about Apple's "cash hoard".


Comments and Sharing


CATEGORIES:




COMMENTS (11 to date)
Philo writes:

"[T]he market for the object being sold is impacted much more strongly than the market for the object being purchased." I don't think this is really what you want to say. When Delta buys an airliner from Boeing, the plane is being sold (by Boeing), but it is also being purchased (by Delta). U.S. money is what is being "purchased" by Boeing and "sold" by Delta, though we don't use those terms for *money*, as opposed for ordinary goods and services. Perhaps your point is that the Delta-Boeing transaction will have more impact on airliner prices than on the value of the dollar because it will constitute a sizeable proportion of annual airliner sales, but only a trivial proportion of transfers of dollars.

Brian Donohue writes:

Great post, Scott.

baconbacon writes:

Well if I had known that Apple had been granted the right to print money I definitely would have bought stock.

If a silver mine opened up and sold silver for dollars AND we measured dollars by the amount of silver they bought- then the sale of silver would effect the dollar the same amount that it effected silver (by definition). The Fed prints dollars and buys securities. We measure the value of securities in terms of dollars.

Some blogger I read has said that monetary base and interest rates are bad indicators of monetary policy- so I wouldn't put to much stock in total purcahses by a CB as a sign of looseness/tightness. On the other hand this other blogger I read has also said "never reason from a price change" so I also would not claim to know why NGDP futures prices rose.

"Central banks are unique because of what they sell---they have a monopoly on base money production---not because of what they buy."

This is true, but why they buy is unique. Unlike investors, central banks aren't concerned with return on investment or risk of loss. This is why I think that they introduce large distortions in the market.

Kevin Erdmann writes:

Is there something I am missing? You seem to be accepting the idea that monetary expansion was related to rising real estate values.

But, the monetary base was contracting from long term trends during the 2000s. Here is a log graph of monetary base since 1959, which is very linear until about 2005, when it begins to flatten.

Here is a graph of Year-over-Year % change in the monetary base, where there is nothing unusual at all until around 2005, when it persistently dips to growth rates below anything we have seen since 1960.

CMA writes:

"But wait (you must be thinking) don't Fed OMPs of Treasury debt cause debt prices to rise? Not always, massive debt purchases in the 1960s and 1970s caused bond prices to fall sharply."

That's because of the income effect. If fed policy generates the same income effect without purchasing bonds through hel-e drops bond prices will always be lower (even in the 60-70's).

CMA writes:

In my previous comment I meant bond prices (or whatever asset is purchased) will always be at a higher level using OMP's than other tools like hel-e's. This is due to the higher demand for the asset by the fed.

Scott Sumner writes:

Philo, Maybe I should have been more specific, I meant exactly what you said in your final remark.

Thanks Brian.

Bacon, You said:

"Some blogger I read has said that monetary base and interest rates are bad indicators of monetary policy- so I wouldn't put to much stock in total purchases by a CB as a sign of looseness/tightness."

Exactly.

You said:

"On the other hand this other blogger I read has also said "never reason from a price change" so I also would not claim to know why NGDP futures prices rose."

Yes, it could be more supply of base money or less demand.

Mark, Their purchases represent the federal government buying back their own debt. If you pay back your mortgage do you worry about the risk of doing so?

Kevin, You said:

"Is there something I am missing? You seem to be accepting the idea that monetary expansion was related to rising real estate values."

No, that's not my view, I mentioned it because others have that view. But that common view is hard to defend, as the debt purchases were too small to have much impact on the credit markets.

CMA, I disagree, the effect is ambiguous regardless of what they buy, as you always have the liquidity effect pushing in the opposite direction from the income/Fisher effects.

Glen writes:

And maybe the money supply had little or nothing to do with the millennial housing bubble (yes, by definition, it was a bubble since it rapidly deflated).

Maybe government policies, implemented mostly but not exclusively through GSEs operating in the mortgage market, distorted credit allocation towards residential real estate. Easy money/easy credit allowed anyone who wanted a house to buy one, irrespective of afforability. Prices responded to demand and homeowners began to believe they were rich and spent accordingly, mostly on mortgage credit. Finally, marginal borrowers defaulted, and VoilĂ ! The bubble popped.

CMA writes:

"CMA, I disagree, the effect is ambiguous regardless of what they buy, as you always have the liquidity effect pushing in the opposite direction from the income/Fisher effects."

Hel-e's will result in lower bond prices than OMP's. If the fed expands M by purchasing bonds it places more upward pressure on bond prices than if it expands money by the same amount without purchasing bonds. Do you disagree with this somehow? I can't see any other possibility.

Scott Sumner writes:

Glen, I agree that monetary policy had little to do with the housing "bubble." Whether there is such a thing as bubbles depends in part on how you define the term 'bubble.' There is no generally accepted definition. Many use the term refer to prices that are clearly well above fundamental values, others to a sharp rise and fall, as you suggested.

CMA, This is a completely separate argument. The effect on interest rates might be different, but I doubt the difference would be significant--perhaps just a basis point. In any case there is never any justification for helicopter drops, which are an exceedingly costly and wasteful policy.

Comments for this entry have been closed
Return to top