Saturday, October 01, 2016

Janet Yellen is right: central banks' purchase of stocks may be useful and justified

When Janet Yellen was being chosen as the boss of the Federal Reserve, I didn't really know who she was, what she knew, and I was somewhat skeptical that she was an extremely bright economist.

Larry Summers – whom I intimately (don't overstate this word, however) know as the former president of Harvard – was an example of a guy whom I often disagree with but whose thinking was expected to be more penetrating, impartial, rational to me than Yellen's. After all, I have known way too many examples in which a less qualified female was picked by the forces of affirmative action.

I must say that after I have watched several press conferences featuring Yellen, I have largely changed my mind. As far as I can say, she understands economics, the economy, and the forces and pressures that affect it. And she is remarkably rational and impartial when it comes to the evaluation of the relevant questions.

During the latest Fed press conference I have watched, she was repeatedly asked about the "politicization of the Fed" questions. Donald Trump has said that due to its loosening and ZIRP policies, the Federal Reserve has been more political than Obama and things like that. You know, I have endorsed Donald Trump as the presidential candidate but I still think that numerous things he says are just silly and untrue and this one is an example.

Yellen said that the U.S. central bankers haven't debated political consequences of their actions; and they haven't debated the influence of partisan politics on their work, either. Minutes that will be released in five years will prove it. I simply believe her. I don't see a reason why I shouldn't believe her. The influences unavoidably act in both directions: monetary policies sometimes affect the competition of the Democrats and the Republicans (or other factions or individuals, I don't want to make the sentence too complicated); and the Democrats and Republicans sometimes try to influence the central bank in order to improve the odds in their competition against the competing party.

However, the U.S. laws – and similarly laws of other Western countries – simply define the central bank as an institution independent from politics and I think that this principle is extremely correct and important. And the Federal Reserve is basically guaranteeing this independence – they decide according to non-partisan rules, whether or not their actions help one party or another. This independence seems real to me despite the obsessive efforts of some politicians, journalists, and pundits to "interpret" all actions of the central bank in a partisan way. At the end of the day, the main job of every central bank is to preserve the "fair value of the money". When someone sells you something for $1,000, he wants to have some near certainty that those $1,000 will be sufficiently valuable at various moments in the future. On the other hand, you don't want the value of those $1,000 to grow, or grow too much, because it would have made your purchase stupid.

Every contract involving the money has two sides. One side wants the price of the money to go up as quickly as possible – because this side has increased its exposure to the money. And the other side – whose exposure to cash has decreased – has the opposite desire. I could talk about loans and many other contracts, too. Clearly, there has to be an entity that is as impartial as possible that determines the right value of the money. It's very important that "borrowers as a group" or "lenders as a group" don't acquire a "larger influence" on the central bank that dictates the value of the money. That's why for American businesses and citizens, it would be much better to use e.g. the Iranian currency than a currency regulated by a union of lenders (or a union of borrowers). The Persian folks would still be better "impartial arbiters".

Now, the value of the U.S. dollar (or other currencies) in the future is approximately predictable because the regulations suggest that the central banks should target the inflation rate – usually around 2%. So in a given year \(Y\) in the future, you may calculate how many "baskets of mixed products" you will be able to buy for your $1,000. In the past, I have defended better choices of the basket and/or the targeting of the nominal GDP growth rate (e.g. around 5%).

There are various choices a central bank can make when it defines the inflation rate or the basket it is based on. Various things may be included in the basket. Food and fuel prices contribute because everyone needs to buy those all the time. On the other hand, they are also more volatile than other things, so they're often removed from the "core inflation rate". Then you have things like clothes and consumer electronics and cars which do contribute to the inflation rate as well as the core inflation rate.

But you could include houses and other real estate to the "basket that measures inflation", too. It makes some sense because most people buy some house or real estate during their life, too. When the house prices go up quickly, "some kind of inflation" is increasing, too. Finally, should you include stocks into the basket that quantifies the inflation? Stocks are something that people often buy and sell, too.

There are many questions but there are no God-given answers. Whenever you try to say that one answer is better than another, you should be able to provide an argument showing that "it will be better for the health, growth rate, and resilience of the economy" if you choose your basket or targeting in one way or another.

I have often promoted the "backing of the currency by a basket of stocks". It's physically impossible to "back" all of the money in circulation by gold. If you assume at least approximately market-driven price ratio of gold to oil, gold to an iPhone etc., gold simply represents a tiny percentage of the mankind's wealth today. So you simply can't "back" everything else – or the money that is prepared to buy everything else – by gold.

On the other hand, stocks are more valuable. The value of each company in the list Apple, Microsoft, Google, Exxon, and others that top the capitalization list is comparable to half a trillion dollars. If you picked thousands of companies, you could get to tens of trillions of dollars of their market capitalization (overall value of all their stocks). The total value of companies is much higher than the total market value of all gold ever mined which is said to be comparable to $8.2 trillion, just some 1/10 of the world's annual GDP.

Central banks simply can't guarantee to pay the holders of all the useful cash the gold of the corresponding amount (by the market price) because there's not enough gold for that. This promise would get broken rather rapidly because the central banks would run out of gold. Let me also mention that the people who want to revive the gold standard after artificially "inflating the gold price" by a factor of 100 or 1,000 are just deluded crackpots. If your plan depends on the distortion of the price of gold (or anything else) by orders of magnitude, then your planned currency is not really covered by gold but by the political power that is needed to force the people to believe that the gold is 100 or 1,000 times more valuable than what the market said before these distortions. If you can fake 99% of the (inflated) gold price, why can't you fake 100%? Forcing the people to believe that one ounce of gold costs $1 million is no different from forcing them to accept that 10,000 pictures of Benjamin Franklin cost $1 million. Of course you can do both. But the presence of "gold" in the "gold standard based on the inflated gold price" doesn't make the system any fairer, more logical, more capitalist, or more resilient than the Benjamin-Fraklin-based counterparts. In both cases, an overwhelming majority of the value or the iPhones you may buy for 1 ounce of the gold (with 100 times inflated price) or for $1 million in banknotes depends on a social convention, not the intrinsic value of the payment medium (gold or pictures of late politicians and scientists).

Because the companies cover such a big part of the mankind's wealth, it's actually sensible to think about the stocks-covered fiat currencies, as a replacement of the obsolete and impossible-to-resuscitate gold standard. It is not hard to figure out that a currency covered by a mixture of stocks has one mathematically equivalent consequence: the corresponding stock exchange index is constant or, with a likely modification, is increasing at a fixed (or predictable) rate every month and every year.

Now, on Thursday, Janet Yellen made a speech in Kansas in which she said that it could be "useful" if her central bank could buy stocks and corporate bonds.

The Feds have apparently discussed this possibility and the only stumbling block is the U.S. Congress that would have to permit such operations.

ECB, in its quantitative easing campaign, is already buying corporate bonds. Banks like the Czech National Bank are holding some 8-10% of their reserves in stocks, mostly U.S. stocks (they attempt to copy the major U.S. indices and a few other indices in the world). It would be much more reasonable for the Czech National Bank to buy stocks traded in Prague, however. Aside from the natural "locality", one reason is that this is a way to make (some) Czechs wealthier – to increase the money supply on the Czech territory (the Prague index is currently below 50% of the pre-2008 record high). Another reason is that when the crown is allowed to strengthen next year, the value of the foreign stocks in the ČNB reserves will drop by the full percentage change of EURCZK; the Czech stocks should drop less than that and some of them don't expect any change.

Janet Yellen says that the "stock purchases" could be useful e.g. during QE4, the fourth wave of quantitative easing. These days, most Western central banks are trying to achieve a much more specific, asymmetric, and down-to-Earth goal: to increase the productive money supply and the inflation rate. The inflation rate near 2% is believed by many economists to be "healthier" than 0% for the growth of the real economy. I don't really believe such stories but I still think that central banks should try to fulfill their commitments.

To print the cash isn't enough. You need to make the cash circulate in the real economy. So as the guy who prints the banknotes, you have to offer the cash to the people for their things. In practice, it was mostly the bonds – the U.S. treasuries etc. – that the central banks were buying. No one wants to establish communism so the cash isn't being thrown from Milton Friedman's helicopters yet. A rule is that the central banks are still buying other things at the current market rates, according to the rules of supply and demand. By increasing the demand, they still push the prices of the corresponding objects up but this is what every market player does and "is allowed to do".

But the central banks have largely run out of the eligible debt – the bonds that are sensible to be bought. So the quantitative easing can't continue for too long, at least not the same kind of quantitative easing: the Feds and the ECB have run out of the gunpowder in this sense. The effect of the quantitative easing programs was arguably nonzero – they increased the amount of the "normal cash" among the people that the people are somewhat more likely to quickly spend or invest (to "real things") than the bonds (although at the end, there is not much difference between the U.S. bonds and U.S. dollar banknotes – they're just "two types of currencies" printed by the U.S. public institutions that simultaneously co-exist and the QE programs were just distorting the relative value of these two co-existing currencies). So the QE programs have "somewhat worked" but the work they have done to return the inflation to the target was "modest" and it's hard to increase it.

I find it obvious that with the stocks, the effect could be stronger, faster, and more logical. The stock prices are more volatile so more money may be pumped into the stock market. And the stocks are also more directly linked to the "fear index". And it's really the "fear index" that the central banks should be taming. Critics ask:
The problem is if the Fed starts to buy stocks, what is the mechanism by which that's going to help the real economy?
Well, it's simple. If a central bank buys stocks in its own country, it means that the values of the stocks will go up. Also, there will be an increased expectation that the growth rate of the stock price will be higher in the future as well because there might be new, institutionalized buyers. It means that everyone who owns stocks – e.g. in the retirement savings – is wealthier right now and, perhaps even more importantly, has better expectations about his future wealth and its growth. That's why he can spend more than he would otherwise spend. The real economy should accelerate as a result.

Also, if the central banks "partly" backed their fiat currencies by stocks – e.g. guaranteed that the major stock indices never drop by more than 20% in a year (by guaranteed purchases from newly printed banknotes) – people would be more certain that a certain kind of a "Great Depression" won't arrive. This would allow them to spend more generously. Needless to say, there is a price that one pays for this remarkable win-win solution. This win-win solution ultimately does increase the spending and the prices grow as a result. The inflation ultimately does grow and these policies have to be stopped and perhaps reverted (the central bank may want to sell the stocks it has previously bought if they become too expensive; or increase the interest rates – this weapon is basically unlimited). But again, I am confident that the inflation is the only "macroeconomic risk" associated with the purchase of the stocks by the central bank.

So far, the Federal Reserve has only been buying government bonds. It has increased their price i.e. reduced the yields. Some part of it must be interpreted as a distortion of the market. In particular, the holders of bonds have benefited from the program more than the holders of the stocks because the bonds were being bought "directly".

In this sense, the quantitative easing programs have created a bubble in the bond market. I think it's a matter of common sense that the central banks' purchases should be as balanced and diluted into many or all asset classes as possible. If the only goal is to increase the money supply or the inflation rate – to reduce the value of the money – they should buy as diverse things for the newly printed cash as possible. Stocks should undoubtedly be a part of it. When central banks buy the "most democratic color-blind mixture" of bonds, stocks, corporate bonds, and other things, you may say that these central banks are not creating bubbles anywhere.

I have presented this way of thinking in many previous blog posts. But the key point is that "one U.S. dollar \(D\)" is just a unit of wealth that is a priori ill-defined – and it's the whole function of time \(D(t)\) that is ill-defined. There is a redundancy – a noncompact version of a \(U(1)\) gauge symmetry – that allows you to rescale all prices by some \(\exp(\lambda(t))\). The commitments and operations done by the Federal Reserve are nothing else than a way to approximately define \(D(t)\), to calibrate the value of the fiat currency, i.e. to gauge-fix the noncompact \(U(1)\) i.e. \(\RR^+\) gauge symmetry.

People who say that this job should be left to the free market don't have an idea what they're talking about. This comment is as nonsensical as the comment that the "numerical value of the speed of light in meters per second" is up to Nature. It just can't be merely up to Nature because it depends on the meaning of "meters" and "seconds" which were introduced by humans. The numerical value of the speed of light in some units primarily depends on the definition of the units which are often man-made (in more natural, Planck, units, we have \(c=1\)). In the same way, the "overall" level of prices in U.S. dollar depends on the definition of the U.S. dollar. NIST is an authority that defines one meter etc.; the Federal Reserve is analogously the authority that defines the U.S. dollar.

Only dimensionless quantities such as price ratios or velocity ratios may be said to be independent of conventions – or independent of institutions that prescribe these conventions.

So my general comment is that the U.S. dollar or other currencies should be basically "backed up by a basket" of almost all things that can be bought. The reserves should be diverse and the central banks should buy everything "equally", in the "fair proportion". In this way, the absolute price level may be modified by the Federal Reserve – and it should be because the Federal Reserve and other central banks are here to guarantee a predictable value of the currencies, to fulfill their commitments that drive the value of the currency.

On the other hand, all this piling of the reserves, quantitative easing, or related programs should be done as finely as possible not to distort the relative prices i.e. price ratios, especially not the price ratios that are determined by the free market in all the healthy situations and the free markets directly decide about "who is a better competitor" or "what is the better allocation of resources and capital".

But I think that a fairer, more stable, and more resilient system would be created if the central banks' portfolios included publicly traded stocks, corporate bonds, and many other kinds of stuff one can buy; and if the central banks were trying to make some index \(I\) basically a predictable increasing function of time – like\[

I(y) = I(2016) \cdot \exp[0.05 (y-2016)].

\] Maybe this identity should be imposed up to some "error margin". If I simplify just a little bit, S&P 500 could become boring, uniformly growing by 5% every year. However, I actually think that it could be a good idea for the central bank to "softly regulate" some industry-wide indices, too.

If the S&P 500 index were prescribed in the way I sketched, you would no longer be afraid of another Black Monday in 1929. Whenever all stocks – according to an index – start to hysterically drop, the central bank is obliged to buy them and compensate for the decrease. Because you know that it would happen in advance, you wouldn't be hysterically selling to start with. The stock market is largely dominated by people who only distinguish an overall good mood and overall bad mood and who buy everything or sell everything (or buy and sell assets according to their predetermined levels of risk). This hysterical mass buying and mass selling is useless for the market which is why the "overall mood" should be kept basically constant by the central banks.

But I think that even industry-wide indices should be made somewhat more stable although not "perfectly stable and predictable".

One should carefully differentiate between the price ratios that are really important and decided by the free market; and those that are similar to the overall value of money and largely arbitrary.

Take the ratio of the Apple-to-Microsoft or Pepsi-to-Coke stock prices. Clearly, you don't want the central banks to distort them or manipulate with them. When Microsoft does something unwise or it's just unsuccessful, it's perfectly sensible and important that the Microsoft stock prices decreases more (or increases less) than the Apple stock does. That's why the central banks should never be picking any winners. They should be buying "neutral baskets" across an industry in a nation.

On the other hand, what is the "right" ratio of the price of a Tesla stock and the stock of Philip Morris? These are two companies that aren't really competing with each other. To make an estimate about the value, you need to make assumptions about their dynamics in the future. Tesla has the huge value because of the speculations that these cars will drive you to Mars in 2050 when millions of Tesla drivers happily live on the red planet. Much of the value of Tesla is about the predictions of a meteoric growth, hopes that are largely supported by the human gullibility, I think, but a few companies like that will surely prove me wrong. On the other hand, Philip Morris' future is about the hope that people won't ever stop smoking even though it seems likely that they are actually reducing smoking.

So the comparison of the stock prices of these two companies – Tesla and Philip Morris – is an extremely subtle exercise. When the Tesla-to-PhilipMorris price ratio dramatically increases or dramatically decreases, it's largely due to some either irrational or "macroeconomic" mood swings and the volatile discount rate implied by them. I think that when the futuristic-car-industry index were "approximately tied" to the cigarette-company index, it wouldn't be bad for anyone.

These two groups of companies have largely nothing to do with each other (unlike pairs of direct competitors, like Apple and Microsoft or Coke and PepsiCo). So if the Federal Reserve guaranteed that the ratio of the sector-wide indices won't change by more than 20 or 30 percent a year (in every 365-day-long period), I think that it wouldn't hurt the efficiency of the market at all. On the other hand, it would increase the stability.

Cigarette companies may fade away and the total capitalization of the cigarette companies may be a 3 times smaller fraction of all companies in 2040 than now. But I don't see a reason why the ratio of "sector indices" should change by more than 20% a year, for example.

So my proposal is to enact a lot of automatic rules that will basically make the central bank buy a lot of "baskets in various sectors" if the index describing the sector was doing really badly in the latest 365 days, e.g. dropped by more than 20%. The central bank would still buy all the companies in the sector according to some fair, proportional rule. Note that 1.22040-2015=95, more than enough to guarantee big swings over the decades.

On the other hand, when some whole sector were increasing by more than 40% a year and the central bank would already own the stocks from this sector in its portfolio, it would be selling these stocks – again, in some proportional way.

Sometimes you might think that your humble correspondent is a potentially "excessively original thinker" while Janet Yellen is a "part of the establishment". But I actually believe that when Janet Yellen were (or is) reading my remarks above, she would (or does) agree that when it comes to the broader philosophy, she's on a similar frequency.

To start with, I do think that the U.S. Congress should allow the Federal Reserve to buy stocks and corporate bonds and the central bank could start to refine its mechanisms to purchase assets in ways that improve the stability but don't produce distortions and bubbles; that reduce the volatility of those ratios whose volatility is harmful while preserving the market-decided volatility of the price ratios that are essential for the allocation of capital and the identification of market winners in each industry. Central bankers should talk about similar mechanisms and the politicians should gradually give them higher permissions to bring them into reality.

I just learned that the previously (almost) unknown 2000 song by Ivan Mládek, The Prague Mosque, has surpassed half a million views, thanks to the recent migration wave – which made Mládek look like a visionary of a sort.

Well, I have been an admirer of his songs and humor since my 4th birthday, of course. He's sometimes tough and the humor is sometimes obscene but I do believe that this song is gentle and remains ambiguous on the question whether he would tolerate a substantial Muslim community in Prague and who is the main target of the satire.


0:17 Prague is very picturesque,
there's both a castle and an observatory there.
Churches and cathedrals,
pubs and shops.

0:25 Bridges, weirs, lock chambers,
exhibition halls, theaters,
but someone is saying
that something is missing here.


0:33 Look at the Czech man, the bighead,
he doesn't want a mosque here.
He is partial to God
won't tolerate Allah here.

Out of Mohammed, the prophet,
he would probably run amok.
I think that out of Islam,
his mouth would probably get sour.

0:50 Look at the Czech man, the bighead,
he doesn't want a mosque in Prague.
He's afraid of [losing] the tripe sausage,
even more so about the 12-degree (5-percent) beer.

He's harming himself, however,
trampling on his happiness.
Out of the hill across the Nusle valley,
a Muslim would be hollering.

1:06 He would be waking us in the morning.
But according to the Quran,
I wouldn't have – which is taboo here
be faithful to only one babe for the whole life.

No comments:

Post a Comment