My Conversation with Larry Summers

Larry was in superb form, and we talked about mentoring, innovation in higher education, monopoly in the American economy, the optimal rate of capital income taxation, philanthropy, Hermann Melville, the benefits of labor unions, Mexico, Russia, and China, Fed undershooting on the inflation target, and Larry’s table tennis adventure in the summer Jewish Olympics. Here is the podcast, video, and transcript.

Here is one excerpt:

SUMMERS: Second, the VIX — people tend to underappreciate this. The volatility of the market moves very much with the level of the market. The reason is that if a company has $100 of debt and $100 of equity, and then the stock market goes up, it’s 50/50 levered.

If the stock market goes up by $100, then it has $100 of debt and $200 of equity and it’s only one-third levered. So when the stock market goes up, its volatility naturally goes down. And the stock market has gone way up over the last 10 months. That’s a factor operating to make its volatility go significantly down.

It’s also the case if you look at surprises. The magnitude of errors in the consensus estimates of company profits or the consensus estimates of industrial production or what have you, numbers have been coming in close to consensus to an unusual degree over the last few months.

I think all those things contribute to the relatively low level of the VIX, but those are more in the way of ex post explanations. If you had told me everything that was going on in the world and asked me to guess where the VIX would be, I would expect it to have been a little higher than it is right now.


COWEN: If there’s an ongoing demand shortfall, as is suggested by many secular stagnation approaches, does that mean monopoly cannot be a major economic problem because that’s from the supply side, and that the supply side constraint isn’t really binding if you think of there as being multiple Lagrangians. Forgive me for getting technical for a moment. Do you see what I’m saying?

SUMMERS: That wouldn’t have been the way I’d have thought about it, Tyler, but what you’re saying might be right. I think I’d be inclined to say that, if there’s more monopoly, there’s more money going to monopoly firms where there’s a low propensity to spend it, both because the firms don’t invest and because the owners of the firms tend to be rich or endowments that have a low propensity to spend.

So the greater monopoly power, to the extent that it exists, is one factor operating to raise savings and reduce investment which contributes to demand shortfalls and secular stagnation.

I also think that there’s likely to be less entry in competition in markets that aren’t growing rapidly than there is in markets that are growing rapidly. There’s a sense in which less demand over time creates its own lack of supply.


COWEN: What mental qualities make for a good table tennis player?

SUMMERS: Judging by my performance, qualities that I do not possess.


SUMMERS: I think a deft wrist, a certain capacity for concentration, and a great deal of practice. While I practiced intensely in the run-up to the activity, there were other participants who had been practicing intensely for decades. And that gave them a substantial advantage.


If you think you know someone who is very smart, Larry is almost certainly smarter.


Way to press him on Russia

"the stock market has gone way up over the last 10 months." -- Good, he agrees that the market shot up when his candidate lost and Trump, who he maligns at every opportunity, won.

Does this make him rethink any of his positions? I kid. He is far too entrenched in the democratic machine to back out now.

The 2017 market looks too much like the 2013 market to draw any big lessons.

Unless you are one of the people who said that a second Obama term or a Trump victory would be the end of the world... (or the demise of America's economy)

It's curious now, because fairly extreme protectionism was on the docket for the Trump administration. The current consensus is that the markets were smart to never believe it? Still not believe it?

"It’s curious now, because fairly extreme protectionism was on the docket for the Trump administration. The current consensus is that the markets were smart to never believe it? Still not believe it?

Too logical question for the tea party (not necessarily conservative, just tea party) commentor above.

Funny how most of the experts think stocks are overvalued. If the download index rises because people are being irrational, is that a good thing?


The "experts" don't know shit though, right?

Smarter than Prof. Williams, a man whose appreciation for his own intelligence is awe inspiring?

In line with the last sentence, listening to many of your podcasts has indeed been quite humbling. Those with Chetty, Heinrich, Pinker, and Rodrik in particular. Even when reading many of these authors, you don't necessarily get the full sense of the scope of their sophistication. It's genuinely awe-inspiring to listen to them riff on such a broad range of topics.

I hate to ask a stupid question, but does a company's leverage change with the market value of its common equity? As opposed to the value of its assets?
And in any event surely a sharp drop would generate a rise in the VIX even if the market level remained much higher than a year ago?

He's saying that if you have some shock that affects the assets of a company by X%, then that shock is going to hit the stock market less hard if the equity/debt split has become more weighted toward equity due to a higher stock market, because the equity slice is the one that has to be flexible while the debt stack doesn't move at all unless the company is in serious trouble. So there's more cushion to absorb that asset shock.

Without having to dig up a bunch of numbers and do a bunch of math, I'm not sure if that makes a ton of sense in this context because most companies just target a debt/EBITDA level anyway, so if the market is moving because earnings expectations are better (rather than just multiple expansion), managements will issue debt or buy back shares as needed to keep everything in line with stock moves.

If I'm not mistaken, I think cjcjc is referring to that fact that debt-to-equity ratios (i.e. leverage) are typically calculated using the book values of equity rather than a market-based values? (Debt is a lien on the firm's assets not on its equity value.) So the leverage of the firm is immune to stock prices, no?

Wikipedia lists many leverage definitions, but says "financial leverage" is "total debt" over "shareholders' equity." So that seems to fit the "plus $100" math.

Yeah I guess I didn't really answer the question.

I don't think there's any great reason to insist on using debt to total book cap as a measure of leverage for a typical industrial company instead of letting the equity portion move with the market. Like I said the most commonly used metric, except for certain financial companies, is going to be debt/EBITDA, so if the company's ability to generate EBITDA has changed and is reflected in the stock price, it should be considered, especially if internally-generated intangibles aren't on the balance sheet.

Presumably, if the firm's equity is increasing in value, it's assets are, also.

Book values and market values are two different things.

Asset and equity prices are linked by net debt. This link is not proportional but absolute. If equity value is increasing, ceteris paribus, asset value is increasing as well, and they're increasing by the same number (not percentage). This fact has been used to explain stock option volatility smirks, which tend to exhibit lower implied volatility at higher strike prices.

A very good interview as always.

On the question of what to do with $100 million in St Louis, I'd try something crazy and see if it could affect Municipality balkanization. The metro area is a treasure trove of information on how to entrench segregation, both racial and economical, with a non-insignificant legislative component. Tiny municipalities that get their budget mostly from people that don't really live in them, and small, independent police departments that would feel at home in the old days, and interesting school districting decisions.

I think the area would only gain value if St Louis county lost at least 70% of its municipalities, but it takes a lot of resources to create such a change.

Not at all a stupid question. Money changes hands among shareholders when the market goes up. The total leverage of companies that compose the market does not change at all. The talk is about "market equity" which, charitably, one could interpret as money invested in the market. This makes little sense though because Summers is initially quoted as saying "if a company has..."

Really smart people are normally able to express themselves more clearly.

In reply to cjcjc

"the total leverage does not change at all". That is to say neither equity nor debt change.

Thanks! I didn't think it sounded quite right.

"Money changes hands among shareholders when the market goes up". And when it goes down.

Yes, and that doesn't change the debt or equity of the companies comprising the market, either.

Trolling Barkley, I see.

"Trolling Barkley"? Huh? Sorry, I am not sure what this is about, Art Deco, although you like to troll me sometimes. Is this a reference to me criticizing Summers here at times over his handling of the Shleifer matter? Indeed, as suggested below, Andrei might be one of those people smarter than the very smart Summers, although his late uncles certainly were his intellectual superiors, as he would readily agree (I knew both of them well, so can suggest them without fear of either of them being dismissed by Tyler, I think ).

Not going to claim to be smarter than Larry, but I have bested him a few times in arguments, with him agreeing that I did, although some of this was done privately. So, while not going to claim being smarter than him, given Tyler's remarks I shall wallow briefly in a bit of implicit egomania, :-).

Oh, I am a much bigger one than you are, "Just sayin'," which is a really stupid fake name to use here, boy.

Kudos to TC for interviewing Summers, but what do people keep saying he's so smart? Apparently Summers got his start in economics from famous relatives. Doesn't that show he's just well connected? Has Summers taken an IQ test like I have (I scored, in my middle age, 120, which is above average; I would have scored higher but it was a UK test with UK pop culture questions).

As for volatility, Summers is wrong ("So when the stock market goes up, its volatility naturally goes down"), as his relationship only is good from 2005 to today, not from 1990 to 2005, see: Coleman, "Applied Investment Theory" (2017) - "Nor does implied equity price volatility – such as the VIX Index – improve forecasts beyond a month ... and fundamentals are of little use either (Engle and Rangel 2008 ). This finding is amplified in the next chart which shows the linkage since 1990 [to 2015] between the S&P 500 Index and the log of its volatility... Consistent with uncertainty in the relationship found by most studies, volatility and price tended to move together during the first half of the period, and then inversely. This makes it obvious why ex post analyses show no consistent relationship between return and risk (Fig. 4.1 ).

"I would have scored higher but it was a UK test with UK pop culture questions)". [SNIP]

So if you'd known who Cliff Richards was, you'd be, what, @ 140?

"MENSA Member, come on down!" ;)

Summers seemed uninformed and talks like he's surrounded by beta sycophants. His answers on China (and Mexico) were atrocious.

Would prefer conversations with the brilliant and very informed Ray Lopez instead.

Summers has a long record of being strong on some details and very poor on big pictures, strong on analysis and weak on synthesis. His demand shortfall analyses are great examples, very attentive to how demand growth slowed without seeing the bigger picture that the US was a low savings country that had been growing its demand faster than its GDP by further and further reducing its savings rate. He rightly sees reduction of the ratio of capital/labor incomes as a potential source of extra demand growth but makes it sound like a bottomless mine when it is of course bounded and cyclical, and by the way already swinging the other way.

I was not particularly impressed by either the questions or answers of this interview.

SUMMERS: I think a variety of the factors holding down investment — the demographic factor, the fact that you can buy an enormous amount of capital for a very low cost, think about my iPhone — all of that I think operates in the direction of meaning that we’re likely to have this phenomenon of low real interest rates and secular stagnation for quite a long time to come.

COWEN (should have been): Does that mean that governments can leverage up, as has particularly been the case in Japan, but also the rest of the OECD on average? We used to have 60% of GDP as a rule of thumb for max government leverage. In an age of structurally low interest rates, does this limit move up? Or put it another way, at the top of the business cycle, a Keynesian model would have an economy running a surplus and paying down debt, for example, in the US with a 77% debt to GDP ratio. Do you think the budget should be in surplus, or has the prudential limit be effectively raised by low interest rates? Is running a deficit still ok, and if so, should we worry about blowing through Ken Rogoff's 90% threshold?

SUMMERS: ... (let's hear a reply).

@Steven Kopits - Nice 'should have been' answer. I think however your Japan figure is a bit wrong in absolute terms though maybe not in terms of trend: Japan is more like the USA in government as percent of GDP in social transfers (i.e. welfare, public pensions): 15% in 1995 (about the same as in the USA) versus 33% for Denmark, Sweden, Finland, and Germany, France, Belgium, Italy not far behind. Greece is like the USA/Japan, very low. I doubt the numbers have made Japan like Sweden, but perhaps due to an aging population the number has gone up a bit in the last 22 years. All of the above from Peter Lindert's excellent two volume treatise on why and how big government is not as bad as people think (and the people love it), though I disagree with his backwards looking logic (I think we can do better with smaller government, better patents, but I just can't prove it like Lindert can prove his case by simply pointing to the historical record: if it ain't broke, it must not need fixing is his logic).

Read my analysis of Japan, on my website with a link to the article in the National Interest.


You are not taking that discredited 90% threshold seriously are you? Or were you expecting Summers to discredit that remark that Tyler would know better than to make?

Barkley -

For many, many years the IMF and other multilateral institutions used 60% debt / GDP as the maximum threshold of prudent fiscal management. Have we abandoned that standard or not?

If your view is that 90% of GDP presents neither drag nor risks to fiscal stability, go ahead and make your case. How high should debt go, in your opinion? Should we emulate Japan and assume interest rates will go to zero, leaving us nothing but residual re-financing risk? How do you want to play it?

If I take your interpretation, I see nothing wrong with cutting corporate taxes to 15% and forgetting about the deficit. That's the implication of your view, isn't it?


I do not have a magic cutoff. 60% was the old limit for European nations to join the euro, and the IMF went along, although there is no evidence that in terms of things happening to a macroeconomy it is that big of a deal. 90% is also not a discontinuity, with Reinhardt and Rogoff ending up with major egg on their faces for falsely claiming it was based on errors in their data files.

What level of debt is good or bad for a nation depends on many things, growth rates, whether the debt is in the nation's currency or a foreign one, interest rates on the debt, and so on. There is no magic number here that fits all.

I thought fixing the errors did not change the conclusion?


It completely removed the discontinuity at 90%. That disappeared. There remained a negative relationship between percent of debt and growth, although given that low growth tends to raise the debt/GDP ratio, the direction of causation on that was ambiguous.

So, Barkley, you know that 90% is wrong but don't have any clear idea of what might be right. And yet S&P just downgraded China on debt worries. That has no impact on growth? Or is that too high, or too low, or you're just clueless? Should the US be running a surplus, or should we blow out the deficit to 91% of GDP to 2027 per the CBO? You don't seem to have any insight at all.

You are the poster child for Trump voters: an alleged expert who has nothing to add to the conversation and takes no responsibility to form and defend an opinion.


We know what is right pretty much. There is no discontinuity. Why should there be? That was what all the whoop about the 90% was. We know that there is a correlation between a higher debt ratio and slower growth, but again, causation is complicated as slowly growing nations end up with high debt ratios. Further important details include what is the extra spending for? Is the government efficient or wasteful? and lots more. Keep in mind, slow growth lowers tax revenues and raises debt ratios. Go look at the list of countries, some with high debt ratios are in bad shape, some are not. And the burden of the debt also depends on interest rates, which is why Japan's super high debt ratio is not a big deal, along with that it is owned mostly by Japanese, with indeed, who owns the debt and what currency is it in also important.

Bottom line is that there simply is no generalization about some particular level of this number that applies across the board, and there certainly is no discontinuity. Got it? You should be able to figure this out, Steven. It is not all that hard.

Age 62 is not old. Experience is under appreciated. Summers' explanation for his sharp mind at age 62 is the explanation I'd give for someone at any age, namely engaging in dialogues with students and colleagues of a younger age, with both Summers and his younger students and colleagues learning from the experience. Of course, it's called the Socratic method for learning. Yet, elite colleges, indeed all colleges, teach primarily using the lecture method. Is it because professors consider themselves too good to engage in Socratic dialogues with students.

Brooksley Born?
Andrei Shleifer?

How does the stock market going up affect the equity of the company? What universe is Larry talking about?

I'm 100% with his VIX comments. A huge part of academic finance is looking at stock prices instead of enterprise values. Stock prices are so influenced by financing decisions.

Here are a few of the properties of stock prices: nearly random walk, fat tails, more downside outliers than upside outliers, volatility clustering, volatility increasing more in response to a downside shock than an upside shock (GJR-Garch). So much of this can be reconsidered if thinking in terms of enterprise values instead and then just consider equity as a call option on the firm value (so from here on out, that's all I'm considering).

Why would stock prices fall by more than they increase? First, consider the elasticity of stock price to firm value. This is given by the delta*firmvalue/stockprice (from elasticity formula). Since firmvalue>stockprice most of the time, this value should be above 1 for reasonable deltas. In other words, a shock to firm values causes a larger shock to stock prices. Further, as the stock price declines, this value will tend to increase, leading to greater stock price declines for each % decline in firm value. It's most certainly an asymmetric effect, which is what is driving the GJR results.

I haven't explained volatility clustering from this model, but if the firm values have a clustering volatility process describing them, then the equity prices would too. I think there's more to think about here.

I think what you guys are missing re Vix is that he's trying to explain why implied volatility ie the cost of insuring against share price moves goes down when prices go up. If you have ever traded vix you know this to be true. The question here is does his explanation work, and I think yes theoretically, but more important are the math of vix calculation and market sentiment.

" There’s a sense in which less demand over time creates its own lack of supply."

Very profound. Almost epitaph material.

The table tennis sojourn was surprising. Summers was and maybe still is a surprisingly good tennis player, I had no idea he dabbled in table tennis.

Long after Tyler's requests for questions I belatedly thought that it would be good to hear another Kenneth Arrow story -- and Summer provided one.

This was not an interview. This was an exercise in status conferral.

I assume myself included in Tyler's less-smart-than-Larry population, although there is one key topic where I might beg to differ:

"What mental qualities make for a good table tennis player?"

I think the question needs to be broadened slightly, to mental and emotional qualities, since the latter are perhaps as important as the former and they can be hard to disentangle. Along those lines:

- Consistent ability to concentrate and focus in the moment, while disregarding distractions;
- Absolute determination, as in the person who will cling to the windowsill the longest when escaping the fire in the building;
- At least a reasonable level of analytical ability, since there is a problem-solving or puzzle aspect to figuring out (or modifying) one's tactics during a match as the opponent tries to do the same;
- Coachability, in terms of technique and training but also match play -- since coaching is now permitted freely throughout a match, and can make a significant difference. Some players will listen and do what you tell them, others just won't;
- Ability to cope positively with stress and even some physical discomfort (at times) while competing, especially in dealing with anxiety and fear of failure or losing (which can be crippling);
- Patience with, or even enthusiasm for intensive technical and physical training that goes on for many years, can be repetitive and just hurts sometimes.

”What is the rate of productivity improvement in table tennis amongst the very best players? Someone like Jan-Ove Waldner or Ding Ning? Are they a lot better than the best players of 20, 30 years ago or just a little better?"

Because of important rule and equipment changes over the years, it's a little hard to disentangle productivity among the players versus changes in technique and styles to optimize under changed conditions. Waldner, for example, developed using the small ball (38mm), speed-glued rubber (use of wet solvents to increase the elasticity of the racket covering), and hidden serves (where he literally concealed the contact with the ball from his opponent, a practice [mostly] no longer allowed). So if you were to imagine a match between Waldner and say, Ma Long at their bests, a non-trivial question is which rules and equipment they would use. The 40mm ball, in particular, has made a significant difference in the style of play and hence physical requirements for top players -- at least in terms of upper-body strength, because it rewards a stronger and faster swing more than the prior ball did.

Now, the corollary question is always, what if Waldner had grown up using the big ball, how would he have done? I have to believe that he would have excelled nonetheless, because he probably would have adapted physically (the Swedes always trained very hard), and his unique abilities and talents would still have mattered. But there has to be a slight question mark about accomplishing the identical level of ultimate achievement. Timo Boll is an example of someone who grew up in one era and prospered as a top professional in another, so perhaps Waldner would have done the same.

There have been a few revolutions in table tennis equipment and technique -- going from hardbats to sponge rackets, the speed glue and hidden serve era, and now the big ball. The first was the most profound, but the others have also caused significant obsolescence for the prior dominant techniques. My model is that it does take some time for the players and training to adapt fully to a change, even a decade or more. For example, Stellan Bengtsson was the breakthrough Swedish world champion in 1971, at the beginnings of the modern sponge racket era. Yet he will tell you that the 1981 version of Stellan would have dominated himself ten years earlier, because of all the technical progress in the intervening time. Likewise, we are approaching two decades into the big ball era, and I think the cumulative adaptations made since 2000 would pose problems for a prior era player who jumped in the DeLorean to enter a major tournament today.

Comments for this post are closed