Thursday, September 10, 2015

Larry Summers vs rate hike

Harvard's President Emeritus Larry Summers has had a blog since April 2015 – the Financial Times and the Washington Post where he writes columns aren't sufficiently intellectually stimulating. In his newest text
Why the Fed must stand still on rates,
he argues that the U.S. central bank should keep the rates at the technical zero during their FOMC meeting on September 17th i.e. next Thursday. The only other possibility that is quoted as likely (current probability is 34% as measured by some prices in the markets) is the rate hike by 0.25%. I think that the difference between them is so tiny and the consequences are so hard to predict that I don't have any strong opinion. But the way of arguing seems interesting to me, anyway, especially because I consider the author a very bright man.

Summers admits that near-zero interest rates are a sign of a pathology; and that the real interest rates are unusually large negative (–1.52%) but he argues against the hike because
  1. two variables, the Goldman Sachs and Chicago Fed "financial conditions" indices, indicate that the conditions remain tight or not too stimulating
  2. the economy is doing badly, like employment etc.
  3. inflation is low and will probably be low
  4. some very bad event, like one in China, may be unlikely but the zero interest rates will be much more helpful in that bad case than the rate hike will be helpful if that bad event doesn't materialize
My view is that the low inflation is the only truly valid argument here – not a terribly ingenious one – and the rest shows that economics is very far from the thinking that would be considered rational and controllable in natural sciences.

Some of the arguments are hocus pocus, as I will argue below, but what I am also disturbed by is the absence of the obvious counterarguments and a balanced attempt to address them. It's hard to get rid of the feeling that this is the kind of an essay that one could always write down in order to rationalize a predetermined conclusion – in this case, the claim that the extremely low interest rates are better.

Aside from those that I have forgotten, I want to make the following points:
  1. the difference between the rate at 0% and 0.25% is extremely tiny
  2. the 0.25% difference is tiny even in peaceful conditions: in direct contradiction to Summers' claims, such a tiny difference would be even more negligible in truly violent conditions
  3. in a basically healthy economy, and I believe it's a right description of the present, the difference between them only helps or hurts selective groups of people or companies while the overall impact is negligible
  4. too much stimulation always creates bogus growth and bubbles but the sustainable long-term growth rate (which may be the same, higher, or lower in the future than it was in the past) can't really be engineered by stimuli
  5. near-zero interest rates are indeed a sign of a pathology
  6. the U.S. unemployment reached the lowest point in 7 years
  7. efforts to tie the Fed decisions to the Goldman Sachs or Chicago Fed variables sound like a very convoluted "theory" whose validity or usefulness hasn't been justified
  8. there are always risks and some people always love fearmongering
First, let me begin with a simple recipe. If the central bank were simply expected to target the inflation rate, say at 2%, there would indeed be a simple reason to keep the interest rates low (or think about negative ones which are problematic for many reasons): the latest U.S. annual inflation rate measured in July 2015 was just 0.2% and the average of this figure over the first 7 months of 2015 is basically zero.

Predictions that significant inflation should have materialized in the U.S. – and elsewhere – by now have turned out to be wrong. I do believe that the rate decisions in an effective and fair system would be more or less automatic and if inflation or similar figures were the key, the decision would probably be "stand still" now.

But now, the criticisms.

The physical importance of the rate hike is small.

First, the difference between 0.00% and 0.25% is really tiny. For three weeks, the financial markets have seen a rather intense wave of volatility. Most stockholders lost about 10% relatively to the peak. Some investors or companies could have made significant changes/reductions in their plans. Daily changes of stock indices above 4% have taken place. Now, we are talking about one of the interest rates that is paid from cash over the whole year and the question is whether it should be higher by 0.00% or 0.25% relatively to now.

The actual, tangible, physical importance of this decision is clearly being overstated by almost all economists. 0.25% is what you lose or gain on stocks in 5 minutes or so. For companies and others, fluctuations of this relative magnitude occur frequently, quickly, and constantly. The very fact that people are brainwashed by this propaganda that the possible 0.25% rate hike is the most important event that everyone must incorporate into his financial planning is extremely pathological. Such details aren't important and it's counterproductive for markets to be programmed to react to such irrelevant things – instead of doing their business, something that distinguishes (or at least should distinguish) everyone from everyone else.

The importance would be even smaller in truly violent conditions

Larry Summers talks about some risks in China. I think that those risks are small – China's authorities seem to possess more than enough tools to stop capital flight and other things – and even if those risks materialized, they may have a limited impact on everyone else. But let's assume, just for the sake of the argument, that some physical global havoc (and not just a global psychological delusion) starts because of China.

What will the importance of the rate hike or its absence be?

Clearly, if we entered even more volatile waters with big drops of the economy etc., the relative importance of 0.25% paid on your cash per year would be even smaller than it is today. Summers' "final argument" says exactly the opposite: the rate hike may have no impact now but it could turn out to be very important in the case of some (unlikely, he admits) terrible global havoc.

But it's exactly the other way around. The correct evaluation may be obtained by comparing the "signal" to "noise". In calm conditions, the signal of magnitude 0.25% may be "measurable" because the noise is relatively low, too. However, in very volatile or noisy conditions, this signal 0.25% gets totally lost – its impact ceases to be measurable – because the noise around is intense.

I actually think that most people will agree with me that this Summers' argument is just mathematically wrong. But in economics and politics, it's often enough to construct a sentence whose grammar is fine and that is written down or pronounced by an authority. That convinces tons of people, doesn't it? But in science, it would be just rubbish.

Small changes like that only "redistribute", don't change the total well-being

I have made this point many times in the past but let me try to make it again, in a concise way. Summers and other economists view the central banks as groups or miraculous shamans who may engineer a much better economy for everyone by carefully adjusting some numbers such as the Federal interest rates.

Well, I view central banks – those who supervise a fiat currency – primarily as arbiters of "justice" for all contracts between the parties that involved, involves, or will involve some monetary compensations etc. In effect, the central banks may strengthen or weaken a currency (immediately or at assorted time scales). Whenever they strengthen the currency relatively to the previous expectations, they help those who have lots of cash (in this currency) or who were promised to be paid fixed amounts of cash in the future; they hurt the opposite side, those who are in debt or those who have promised to pay money in the future etc.

When the currency is weakened, the fates of the two groups are interchanged.

The point is that this is basically a zero-sum game. As long as the economy is basically balanced, optimized to the conditions that existed before the decision, it minimizes the action (a quantity that measures the deviation of the free markets' adjustments away from some optimum point) and we have \(\delta S = 0\). So if you move the variables by \(\delta x \neq 0\), the action won't change by terms \(O(\delta x)\) at all: it only changes by the negligible terms of order \(O(\delta x^2) \).

Any decision about minor changes such as minor changes of the Fed rates is therefore a matter of redistribution. Now, I am convinced that the central banks shouldn't think of themselves as some kind of oracles or the "ultimate source of life" in the economy – which they are not. Instead, they should try to be as just as possible.

Imagine that Mr Smith buys a villa in Hollywood from Mr Brown. They think about the transaction. After all, they barely decide it's an OK deal and Mr Smith pays $10 million. (This is a template for absolutely any transaction – job offer or anything like that – involving some money.) Why was it the right amount? It's because both Mr Smith and Mr Brown have some idea what $10 million will be worth tomorrow or next year or in 5 years from now. Mr Smith will be missing those funds, Mr Brown will have the extra money.

So there should exist some rules, at least approximate ones – like inflation targeting or something playing a similar role – that allows both Mr Brown and Mr Smith to think about the a priori abstract money with some precision. If the central bank suddenly makes an unexpected move and weakens the currency, or increases the prices of houses, it basically helps Mr Smith who already possesses the new house while it hurts Mr Brown who has lots of cash (whose value went down).

But the actual wealth in the economy is being created – and new houses are being built – by the real people and companies because they become parties in numerous contracts that are "good for both sides". Such contracts may be signed regardless of the strength of the U.S. dollar or the expected interest rate or inflation rate. Both Mr Brown and Mr Smith – and billions of their counterparts – may always take the promises and commitments and expectations about the currency, the inflation rates, and other rates into account. What is important is that the central bank doesn't systematically make the life easier for the likes of Mr Smith – or for the likes of Mr Brown.

Bogus growth is easy to create by central decisions, sustainable one is nearly impossible

If the financial conditions are too loose, it becomes easier to borrow the money, produce products, or build new factories, among other things. And it becomes easier to buy stocks or other risky investments. However, when this happens, it's not guaranteed that the investment was wise. One isn't guaranteed to sell the products. One isn't guaranteed that the new factory will be utilized for the years or decades to come. And one isn't guaranteed that the risky investments will yield great returns, either.

The loose conditions may very well be creating just bubbles, something that won't last forever, something that will burst.

A problem with the GDP figures is that it is simply not possible to predict what the market capitalization or utilization of a company will be 5 years from now but the GDP is already measured now. So the GDP contains both entries that are sustainable and those that are not. The GDP contains lots of activity that is bubble-like. Lots of activity run by loans that won't be repaid because the financial plan isn't sound in the long run.

I think it's obvious that if one tries to pump up the GDP growth rate by more loose financial conditions etc. – in the ways that Summers seems to share as well – one primarily creates the "easy growth" of the second kind, i.e. the bubbles that won't be sustainable and the activity that will contribute to the debt that won't be repaid. One may indeed increase the GDP by making the conditions even more loose, or loose for a longer time than expected, but such a higher GDP figure may be a flawed achievement to celebrate because it's largely fabricated. The bogus activity will turn out to be useless or worthless and it will be subtracted in the future again.

Some hardcore Keynesians are proud about their not paying attention to the difference between the useful growth and the artificially engineered one at all. Paul Krugman wants to break all the windows – or simulate an invasion of extraterrestrial aliens – in order to improve the GDP figures. But he completely overlooks – and, apparently, deliberately overlooks – the fact that the new windows that have to be paid for (because they were destroyed by someone, either a vandal or a fake extraterrestrial alien: the latter – Mr Krugman himself – is a special, institutionalized case of the former) are unnecessary expenses. They don't really make the nation as a whole more happy. If the GDP is meant to quantify the overall material happiness of the nation, it's obvious that the incorporation of the fees for "unnecessarily broken window" is a fabrication.

At the end, there is no simple recipe to guarantee a higher yet sustainable and "real" growth rate by some central decisions. The actual values are not being created in the central bank or by the government.

We don't know what the average real growth rate in the Western economies will be between 2015 and 2065. But it's some number whose value is largely unavoidable. One may build useless new factories – they probably have lots of them in China – but those won't really increase the long-term growth rate because their uselessness or worthlessness will be seen much more quickly than in 50 years.

Just like the central bank can't engineer the real long-term interest rates, it can't engineer the real long-term GDP growth rates. Those numbers are dictated by the real markets and the laws of physics. The central banks and governments may only fabricate numbers that make the situation look temporarily different than it fundamentally is.

In the 19th century, growth rates around 10 percent a year were common for whole decades. But the industrialization was more intense than it is today. Relatively to the changes of their lives that the citizens of the Western countries experienced in the 19th century, the ongoing changes and new models of the iPhones are only changing some irrelevant details.

Also, people may very well realize that they're already rich enough. One needs a certain amount of wealth to live happily. Above this value, the returns are quickly diminishing. So in the future, the real growth rates may be lower because people just don't need much extra growth. They're already satisfied. They produce enough. They may also find out that they don't need to work as hard as they did before. Or they don't need to work 5 days a week. Maybe they want to use the increases of productivity by adding Monday to the weekend. And, 20 years later, also Tuesday. And so on.

If that's so, the GDP real growth figures may be lower in the future but that's how things should be – because that "stagnation" will express the people's evolving opinions about the value of money relatively to the value of leisure time. If that's so and the growth rates will be lower in the next 20 years for this reason, people shouldn't try to be hysterical about it. One may talk about the secular (=not just temporary) stagnation and these words may sound scary but the substance hiding behind the words may still reflect real people's genuine wishes.

Whether Larry Summers or Paul Krugman like it or not, citizens of free countries aren't "obliged" to fulfill some arbitrary 5-year or other plans dictating some growth rate figures. Free people are generally acting to optimize their happiness or well-being. This quantity is rather strongly correlated with the GDP but if you start to "push" GDP by artificial tricks, the correlation between GDP and happiness weakens and it should be the happiness, and not the fabricated GDP, that matters.

Near-zero rates are a symptom of a pathology

The interest rates, and even the real ones (with the inflation rate subtracted), should normally be positive. It should be hard to have "enough money for everything we need now" right away, and it should generally cost money to borrow. The interest rates on a particular loan must combine the inflation rate, the discount rate, the risk of default, and more. Negative rates basically mean that we don't like to enjoy material things now – they're worse than nothing. And technically, negative interest rates make people store banknotes in their mattresses which is bad for obvious reasons and so on. The interest rates were brought to the technical zero to fight against the big depression and its consequences.

One recent event indicating that the markets began to appreciate how pathological the near-zero rates are was a drop of the stock market sometimes in August that followed the announcement that the Fed wasn't decided to hike the rates, after all. Although such a dovish change of the policymakers' attitudes should normally be supportive for the stocks, the reaction was the opposite one. Maybe, the traders were scared of the pathology that was admitted by the officials. Maybe, they became afraid of some "scary dirty secret" that only the Fed insiders know. I think that no such secret exists.

Let us return to the historical reason why we're in the zero interest rate epoch. It started by the 2007-2009 downturn and this downturn was the basis of all justifications for the continuing policy afterwards.

But one should realize that those events – the collapse of the Lehman Brothers etc. – are a history. Since that time, lots of things have happened. Many stock indices have tripled since the 2009 lows. GDP growth rates are rather high in many countries, people feel materially OK. And as I wrote above, the unemployment rate in the U.S. is lower than at any moment in the recent 7 years. It is at 5.1 percent. Given some percentage of people who are very hard to employ and the need of the economy to have some "buffer", I think it's fair to call this level a de facto full employment.

The economy isn't creating too many jobs but it's questionable whether it "should" or whether we should expect it to create them. An increasing number of people may be doing fine without too much work, or any work, for that matter. So the number of jobs doesn't have to rise too much. People may feel rich enough with the jobs that exist. It's really the traditional unemployment rate that quantifies the under-utilization of the labor force – the preferred quantity in Europe – and this rate is extremely low in the U.S.

All the claims that the U.S. or world economy is still very weak and facing lots of diseases are fabrications. Whether such comments are right depends on where you place the reference point. But making too optimistic assumptions about the economy may be silly. There is no good evidence for the belief that the growth rates etc. will be substantially better e.g. in 2018 than they are now.

Instead, I think it's fair to say that 2015 seems to be one of the healthiest years in the economy over the recent decades – and yes, one should really evaluate the quality of the economy by comparisons to other years, not by comparisons to some arbitrary dreams, centralist plans, and utopias (which is what Summers seems to be doing). To present the conditions as a reason for unusually loose policies seems utterly demagogic to me.

Convoluted indices as the criteria for policies

Larry Summers offered two indices, one by the Chicago Fed and one by Goldman Sachs, to argue that despite the clearly negative real interest rates, zero nominal interest rates etc., the financial conditions aren't too loose in comparison with the recent 3 decades.

Great, fine. Some 2 institutions have some variable and it may be picked to make a point. But what's the overall logic of this point? Summers basically claims
Because the Chicago Fed financial condition index is above 0.4 and the Goldman Sachs one is above 100.7, indicating "conditions that aren't too loose", the central bank should keep zero interest rates.
Now, what is the justification of such a claim, except that it's a hocus pocus rule-of-thumb that may sound convincing to someone who doesn't think critically?

The rule above isn't any law. It is not a part of the Fed's policies or regulations, either. It's a new proposed "law of physics" that was invented by Larry Summers. If he wants to say that it would be wise for the Fed to decide about the interest rates according to the Goldman Sachs and Chicago Fed indices, he would need to be more specific about what the rules should be; and why these rules are claimed to be good for the economy or anything.

He doesn't have either. So I think it's fair to say that he's just using two random quantities to present an argument to keep a third quantity at zero – even though he doesn't have any solid argument or theoretical basis for such constructions.

Even the inflation targeting may look rather complex to many people. But the "financial conditions" indices are even more complex and more abstract. If the Fed's decisions were dictated by such obscure quantities, the policymaking would basically become an unreadable and unpredictable shamanism, I believe.

Persistent fearmongering

Summers talks about the risks associated with China. How big they actually are? Are there any reasons to believe that they differ from similar risks that people were afraid of at almost every moment of the history (except for moments when catastrophes were actually around the corner)?

At the end, it seems hard to avoid the feeling that Summers only collects arguments and pseudoarguments that support one possible answer to the policy questions – the answer that the more loose conditions are always the better choice.


I think that after some time, perhaps 2 years, perhaps 5 years, the loose conditions will create inflation. When people face negative real interest rates for a long enough time, they will be increasingly storing their savings as well as "more short-term cash" in stocks. They will adapt to the new situation in some way. Increasingly loose financial conditions may reduce the value of cash – but people unavoidably find the other ways to store wealth which don't lose the value. So the idea that the government may and should force people to spend more – and reduce the spending rate – is both "going against the people's freedom and their actual wishes" and "impossible to realize in practice".

Numbers such as the nominal GDP growth rate may mean different things than they did in the past. The more convoluted justifications for unusual policies are used, the harder it will be to extract useful information from the numbers that used to be rather useful, and the closer to voodoo the policymaking of the central banks may become.

In some cases, "less is more" and more straightforward arguments and quantities (such as the observed inflation rate) may be better guides for the monetary policymaking than some excessively contrived – and unverified – ones.

No comments:

Post a Comment