Sunday, 25 June 2017

A change of plan? Random thoughts prompted by Tony Yates.

Tony Yates wrote a post on his blog on Friday- Balance sheet shrinkage: so soon?  - asking why the Fed is considering reducing the size of its balance sheet, while interest rates are still very low. People who spend their days staring at screens and messing around in financial market impact tend to think more about what policy-makers will do and what it will mean for markets, than about why they are doing something. Maybe shrinking the balance sheet is the lesser of two evils (the other being to go on increasing interest rates at the current pace in the face of low inflation. Or perhaps they just think the markets are in a good state to cope with them acting now, and it's just a pragmatic decision. But it does suggest that the Fed, at least, is still uncertain about the long-term implications of carrying a lot of assets on its balance sheet.

In the beginning, central banks' first steps into the murky waters of 'QE' were tentative. They clearly wanted to buy as much as necessary but as little as possible, frightened of long-term side-effects that they couldn't yet contemplate. The same was true of negative interest rates. But as time has gone on, we and they (the Fed et al) have started feel that we understand the impact of central bank intervention in bond markets little more. And yes, a little knowledge really can be very dangerous!

When the Fed first bought bonds, there was a flurry of research papers trying to gauge how big the impact would be on bond yields, particularly he yield on longer-dated Treasuries. Disentangling the relative importance of the bond-buying from the accompanying (super-low) interest rates, along with the longer-term economic consequences of the financial crisis (modest growth in GDP, low wage growth and below-target inflation inter alia) was never going to be straight-forward.  But it's been easier for the sell-side research community to get its head around what bond-buying meant for the yield spread between government debt and the other bonds that investors were being crowded into.  These three charts, posted on twitter over the last few weeks from research reports written by BofA Merrill Lynch and Citi's Matt King, sum up the state of thinking.

 1) The net supply of bonds to the private sector investor universe has dwindled to almost nothing, if central bank buying is subtracted. A disaster if you are trying to get some decent yield into your pension portfolio, a recipe for investors to go hunting in ever more exotic places in their search for yield. Hundred-year Argentina bonds anyone?

2)  A similar story from Matt King -  if the Fed stats selling, global bond issuance goes from just below zero in 2016, and a few hundred billion dollars in 2017, to $1.5trn in 2018.  And in the third picture, which is a blow-up of the top-right chart in the second, you can see the conclusion he reaches: less buying of bonds by the Fed et al is bad for corporate bonds and implies wider spreads.

Note that these aren't just charts of Fed bond-buying. They show global asset purchases, and maybe that too plays a part in Fed thinking: If what markets are sensitive to is the collective actions of central banks, maybe it's better to be first to shed excess assets, rather than waiting for the ECB, BOJ and Bank of England to join in, by which time any impact on credit spreads (and on absolute yields) will already be being felt and Fed action would exacerbate it.

I also think it's interesting to consider that the actions of the Fed affect markets differently than the action of other central banks. The next chart highlights the challenge facing the ECB and comes from their latest Economic Bulletin showing the impact of ECB bond-buying on various investor communities.

The ECB estimates that the Eurosystem owns EUR 1.1trn more government bonds than they would have otherwise, as a result of activities between March 2015 and march 2017. A little more than half of that has come as a result of crowding out European banks and investors, and a little under half has come as a result of crowding out foreign investors. It doesn't take a huge leap of faith to conclude that where the main impact of goal bond-buying globally (and Fed buying in particular) has been on credit spreads, the impact of the ECB being has also been felt in the currency market, as foreign investors were squeezed out of euro-denominated  assets (and Europeans were squeezed out of euros, for that matter).

Where does that leave me, as I ponder Tony's question - has the Fed changed its plan and if so, why? Tony's implicit conclusion is that the Fed has changed its plan and he suggests hostility from politicians as a possible reason. Two more spring to mind if I think the Fed is, in part, simply being pragmatic.

Firstly, they might have changed course because the possible effects of running down the balance sheet - somewhat higher longer-dated yields, wider credit spreads, and maybe a knock-on to equity markets -  would suit them. In a world where they face criticism for raising rates at the same time as they make little or no progress towards hitting their inflation target, maybe it's easier to shift attention slightly to the balance sheet. They always wanted to normalise it, and there's reasonable evidence that buying bonds has distorted asset markets. So why not do so that at a time when credit spreads are tight, global bond market are buoyant and the equity market is making new highs? Secondly, they'd much rather be the first major central bank to run down its holdings, than the last. Better not to sell into a crowded market.

The ECB faces a different problem. Concerns about the euro rising as they back scale bank bond-buying, just as it fell when they instigated the programme, seem hard to get around unless they just opt not to taper, and not to even think about thinking about raising rates.

Monday, 17 April 2017

Ben Bernanke and the case for lower for even longer

Michael Kiley and John Roberts, two Federal reserve economists,  have produced a beautiful paper on Monetary policy in a low-interest rate world which in turn prompted ex Fed Chairman (and easy policy champion par excellence) Ben Bernanke to write not one but two articles in Brookings last week:  How big a problem is the zero lower bound on interest rates? and The zero lower bound on interest rates: How should the Fed respond?

Mr Bernanke's man points can be summed up with a few quotes:

Using econometric models to simulate the performance of the U.S. economy, Kiley and Roberts (KR) find that, under certain assumptions, in the future short-term interest rates could be at zero as much as 30 to 40 percent of the time, hobbling the ability of the Fed to ease monetary policy when needed. As a result, their simulations predict that future economic performance will be poor on average, with inflation well below the Fed’s 2 percent target and output below its potential.

One potential solution to this problem, suggested by leading economists like Olivier Blanchard, is for the Fed to raise its inflation target.....There are, however, some reasons that raising the Fed’s inflation target might not be such a good idea....Another alternative would be to try to implement the optimal “make-up” strategy, in which the Fed commits to compensating for the effects of the ZLB by holding rates low for a time after the ZLB no longer binds, with the length of the make-up period explicitly depending on the severity of the ZLB episode. KR consider several policies of this type and show in their simulations that such policies reduce the frequency of ZLB episodes and largely eliminate their costs, while keeping average inflation close to 2 percent. 

The basic premise of the Kiley/Roberts paper is that if the 'neutral' real interest rate is as low as 1% (which is broadly what most analyses conclude), then the neutral nominal rate with a 2% inflation target is 3%. And if neutral rates in minimal terms are at 3%, rather than, say 6%, then the chances that rates ought to be below zero at some pint in the economic cycle are high. which, in turn, causes a huge problem if the zero lower bound makes getting inflation back up to target in economic downturn harder. And that means that there's a bias, with a risk that inflation will be below target more frequently and output below potential more frequently, risking a vicious spiral as the central bank fails to get output back on trend. Hence the Bernanke solution - a deliberate policy of keeping policy to easy into an economic upswing, temporarily ignoring upside inflationary risks. 

This is interesting because it seems (to me, anyway) that this is exactly what current fed policy looks like. Old-fashioned measures of a potential output suggest that with the economy at or close to full employment (the weakness of wage growth notwithstanding), there isn't much slack left. So rates 'ought' to be headed to neutral (3% in nominal terms) pretty briskly. And yet, while the FOMC 'dot-plot' does get to 3% 'in the longer run',  insouciance at the market pricing which undershoots the dots, and the tone of Fed commentary, both suggest that there is a deliberate bias not to rush. 
Measures of neutral real rates (mind you, however we look at them we're a long way below neutral). 

I have one fundamental issue with the whole argument however, and that's the idea that potential output is in some way related to inflation in the shortish-term. Downward-pressure on inflation from globalisation and more recently from technology,  means that 'potential output' needs to be thought of in terms that go beyond 'non-inflationary output'. 

This means that if measures of 'neutral' rates are trying to be consistent with steady low inflation, they will be inconsistent with wider stability, in asset prices, in the balance of payments and in debt levels in particular. Of course, that brings me back to Claudio Borio and the importance of a financial stability-oriented monetary policy framework. 

Two slightly more market-related conclusions are firstly that despite the BIS' work, and in contrast to other central banks, the Fed still has a bias deep within its core, to favour 'lower for longer' policies, and secondly  that the dollar's previous major bull markets have come when real rates in the US have been well-known above all measures of r*. To say that's not the case at the moment would be a huge under-statement. 


Saturday, 1 April 2017

The rise and rise of debt, correlation and commotion

The rise and rise of debt, correlation and commotion

Trilemma/dilemma + The Second Machine Age  + Financial Sector Drag vs Secular Stagnation  =

If economics is perceived as a dismal science then the onus on economists is to make it less so.
If I could only persuade my teenage children to read one book, one magazine article and one slide-show of a speech in order to pique their interest in the subject, I'd have them read the Economist on Helen Rey's Trilemma, look through the slides of Claudio Borio's speech to the NABE a few weeks ago, and then settle down to read Brynjolfsson and Mcafee's "The Second machine Age" over the Easter holidays.

The chart captures the whole of my career, which started after I left university in 1984. Lower and lower real interest rates, higher and higher debt levels.

Professor Rey argues, convincingly, that in today's connected global economy, if you have relatively free capital flows you can't have monetary independence whether you let you exchange rate float freely or not.  She observes, in the process, that asset price movements are correlate and are to a significant degree a function of the policies of the US Federal Reserve. Put it another way - Fed policy drives all markets, and increases asset correlation as well as leading us to a world where risk is either 'on' or 'off'.

Claudio Borio makes the simple observation that monetary policy which is consistent with equilibrium in the real economy but not in the financial sector, isn't an equilibrium interest rate at all. This is so obviously true it took a genius to point it out. If interest rate-setting central banks are only judged by a mandate of an inflation target and an economy at full employment,  they may have a tendency to set rates at levels which allow the kind of ever-upward march in debt levels seen in the chart. If they took the idea of financial stability seriously, they would set policy differently, but it's not directly part of their mandate and as we can all see at the moment, fear of too-low consumer price inflation is still delivering interest rats that risk increasing global debt.

The Second Machine Age is just a great book in its own right. But it does help explain why low unemployment might not be associated with wage-induced inflation and indeed why inflation might stay low regardless of where interest rates are set by the Fed, the ECB and any other central bank.

If technology is the main driver of subdued wage growth and inflation; if US interest rates are set at levels to keep inflation down when it's down anyway; and if the rest of us import Fed policy through the global finance system, then we can just get used to asset price volatility being depressed, asset prices being hyper-correlated, debt levels going up and the profit share of GDP in most countries. marching slowly higher. All with sporadic episodes when markets that can't adequately price risk are exposed to sudden re-pricing on (often politically-inspired) surprises.

Sunday, 19 March 2017

Meetings - why do we have  to have the stupid things given that they are almost universally perceived as one of the banes of working life?  The best article I've read about meetings was by written by Antony Jay for the Harvard Business Review in 1976 and he kicks off with the comment that "great many meetings waste a great deal of everyone’s time....long-established committees are little more than memorials to dead problems". 

Meetings can be divided into three broad types. 'Town Hall' meetings, 'Cabinet' meetings and 'Pub' meetings. The Town Mall meeting is a gathering  for leaders/bosses or others to impart information and can be as big as you like. Cabinet meetings are part theatre as little is denied but they are there to reinforce the importance of the members of the cabinet and to ensure that everyone its been by each other to sign off on an agreed plan. And a Pub meeting is a place for creative debate that rarely takes place in a pub at all. 

The Town Hall 

The boss wants to give the troops an update on the firm's performance over the last year, outline the key takeaways from the senior management team's offsite in Monaco, and send some motivational messages to help encourage everyone to work even harder in the months ahead. He or she instructs someone to prepare a powerpoint presentation with the  financial results on it, the key goals, some metrics on profitability, on the trend in the workforce, perhaps on the trend in compensation. The boss then stands at the front of the room,  goes through the slides,  and opens the floor to questions at which point a few brave souls ask a few bland questions, depending on part on how many people there are listening. 

It's possible to do these well, but it's more common to do them very badly. In the best case, the speaker is a motivational, charismatic leader  who doesn't need a powerpoint to help sell a vision of the future. If there are slides they are for reference on technical issues. The listener leaves impressed by the vision and by the passion of the visionary. 

In the worst case, the speaker hasn't really read through the slides properly in advance and so is left mumbling his or her way through them, reading off the screen. Perhaps the VC equipment wasn't tested much in advance and a technician needs to be summoned early in the presentation. There is little ad-libbing and so even less passion or motivation. There may be questions but the overall sense as the listeners leave, is of deflation. A ritual has been accomplished, because it is a ritual obligation of management to talk to the common workers occasionally, but little has been achieved and an opportunity has been missed. 

The cabinet meeting

A cabinet or committee meeting is a social structure, where a group gets together to formally reach agreement, and where the members can be seen to be reaching agreement. The group shares information which others may not have, reasserting its own exclusivity. It may have the meeting at a time and in a place where other lesser mortals know it's happening, which boosts prestige. It offers a chance for the members to reinforce their status relative to each other. But the key to these meetings is that they can only reach significant agreement if there has been sufficient preparation. Issues have been debated in smaller groups over a period of time and the purpose of the final gathering is to iron out the last wrinkles, put the finishing touches on the plan and to be seen to be in collective agreement so that there is collective accountability. No-one can easily say they were never part of the agreement.

So it's a meeting to agree something that's already been agreed.  I'm sure there are some things that are agreed at a cabinet meeting but the preparation work has been done in advance. A proposal was made, the meeting participants were canvased, their views heard. The t's are crossed and the i's are dotted at the meeting itself but the purpose is for everyone to see the body-language, and then there's photo-opportunity, or a signing ceremony or some sort of mutual affirmation.

The failure of a vast number of meetings in real life, is that they should be this kind, where the groundwork has been done long in advance, but instead there is some absurd hope that a group of 20-odd people can get together, discuss, debate and agree something from a standing start in an hour-long meeting. It's impossible for a host of reasons but the most important is that there's so much ritual involved in this kind of meeting. Who sits where, who speaks when, who is trying to impress the boss, who is competing with whom to be a bigger cheese week than they were last, and so on.  If you chair meetings of 6 to 20 people in a work environment and want to understand a bit of the underlying psychology, try asking the members whether they mind the meeting being filmed for training purposes and stick a few discrete cameras in the corners of the room. There will be people trying to get a word in edgeways who are ignored. There will be people whose sole purpose is to disagree with their competitors. There will be body-language galore as our inner ape goes into overdrive.

I've sat through some good meetings on business planning, some meetings where the homework that had been done in advance really paid off and strategic plans were proposed, debated from the perspective off everyone having already done a lot of thinking abut them, and agreement reached. I've sat in more and ones were someone was mad enough to think that 20 people could brainstorm a subject and come up with an interesting plan in an hour. Some of the most awful of these meetings, where presenters drone on, anyone who asks a question is scowled at because that makes the meeting drag on for even longer and where the only thing that is certain is that the group will not take genuine collective ownership of the conclusions, can be found in investment forecasting and strategy. A better recipe for half-hearted groupthink is hard to imagine. 

The Pub meeting 

Which is why I prefer the 'Pub' meeting, where issues really are debated and a relatively blank canvas is covered with ideas and eventually a plan. The image in my mind of a pub argument is of people jabbing fingers in each others faces, sometimes disagreeing loudly, sometimes agreeing joyously, always walking out arm in arm. 

In practice, these meetings can only happen in pubs if all the members of the group are the kind of people comfortable in a pub.  And I'm not sure that you can have even as many as ten people participating. But wherever the meeting happens the rules of engagement are the same - everyone is entitled to their opinion and is encouraged to have a say, regardless of status, age, experience, sex, ethnic background or which rugby team they support. Disagreement is encouraged. But finally, it is absolutely mandatory that everyone leave the pub able to put aside the differences of opinion and be civil to each other. 

In an office environment, the best way to run these meetings is to make the older/more senior/more vocal members of the team wait a while before entering the argument. Younger or more junior people will be nervous of disagreeing later, so need encouraging to express views. And need to know in advance their opinions are going to be sought. Those people who in a film of a meeting are quietly trying to get a word in, need to be picked out early too. It's also important to get to the most important issue for discussion early, before the group starts to flag. 

Some people will never be able to change their opinions even in an environment that encourage them to question them. That's fine, so long as they can debate without being (too) boring and aren't allowed to dominate proceedings. The debate is there to dig into the pros and cons of different opinions and also, to tease out relevant pearls of wisdom from unusual sources. The geeky end of the asset-backed research group had all the clues to impending disaster in 2007 but few firms were listening. The premise of Moneyball is the statisticians had more answers to baseball performance than experienced scouts. A good meeting creates a safe space where the views of these people can get a hearing. 

Saturday, 11 March 2017

Debt will be the death of us

As investment bank research teams ponder how much to charge for their insights, they could do worse than check out what is available for free, particularly from the growing band of central bank blogs of which the BIS' is still probably the best. Claudio Borio's speech to the NABE conference in Washington DC last week is fascinating (and free, of course).

The speech looks at the role of financial sector drag as a reason for the sluggishness of the economic recovery in many developed economies after the 2008 financial crisis. The BIS view of the world could be summed up as being that equilibrium or natural real interest rates are higher than many policy-makes believe, and this has resulted in an inability to restrain financial booms which cause a misallocation of resources that in turn, leads to painfully slow productivity gains in the subsequent economic recovery.

This is so intuitively true that it needs to be taken seriously. I've reproduced a couple of slides from this presentation and one from an earlier presentation covering the same theme. The first one shows global debt levels rising as a share of GDP, at the same time as real interest rates fall.

If I made the really simple assumption that across the economy as a whole, borrowers pay something like 3% more than the average of the real yields in long-dated government bonds and the central bank policy rate, and that inflation was around 3% in 1986 and 2% in 2015, this chart would suggest that in 1986, overall debt service costs were something like 19% GDP. And in 2015, after almost 30 years of rising debt and falling rates, they were something like 16% GDP. So, no problem then?

You could argue that it makes sense in this world, for policy-makers to accept the downward trend in interest rates. Companies and households alike would and indeed should have more debt because debt-service costs are low. Which is what's happened. Mortgages are bigger as a multiple of income than they used to be (fine if house prices rise for ever, right?) and a glance at the universe of borrowers in the corporate bond market will quickly confirm the gradual downward drift in average credit ratings. A triple-A company with little debt is not maximising the return to shareholders. Better to issue bonds and buy shares back, immediately.

The flaw, is that creditworthiness is a function not solely (or even mainly) of the ability to pay back the interest on a loan. It's about the ability to pay back interest AND principal. In other words, the size of the loan matters. Furthermore, it matters at least as much for the financial sector as for the non-financial sector. If I borrow money from the bank, the bank itself borrows that money too. And if we've learnt nothing else in the last decade, it's that that size of a bank's balance sheet matters. A bank that cares only about keeping income (from interest earned) growing faster than costs (from interest paid) will grow and grow it's balance sheet much as some did in the early 2000s. A focus on return on equity may be Ok for some companies, but not for a bank which needs to care about the return on the capital it uses to earn that money, and needs to set capital aside for the possibility that some of the money it lends never comes back.

 Just accepting that premise implies that the 'equilibrium' which sees more debt and lower rates, is an unstable one. A real equilibrium is one which applies to both the financial cycle and the real economy cycle. As in the second chart I've nicked.....
 This chart shows that real policy rates fell sharply and stayed low in 2001-2204. This is after the bubble burst and after the 9/11 attacks on the US.  Fed Chairman Alan Greenspan worried about disinflation even as the global economy started to recover in earnest and so the Us kept rates down (and the rest of us followed).  In part because inflation was was falling more than people expected, estimates of the natural rate of interest were coming down, albeit not nearly enough to justify the Greenspan-led monetary madness unleashed on us all.  But adjust them for the financial cycle and the natural rate was already rising in 2002-2003.

The crisis followed. Too much money was lent at these low rates. Too many resources were mis-allocated in a world where money was mis-priced. Here's what the BIS' experiment in setting policy to smooth the financial cycle throws out. A cycle, just not nearly such a pronounced one.

After the financial crisis and recession, the misallocation of resources that the financial boom caused/exacerbated, has led to a productivity-free economic recovery, at least in the economies most directly affected. So, estimates of natural rates have come down further. And since there's slack in the global economy, especially if we measure it on the basis of inflation undershooting expectations, policy rates fall even further. The possibility that the inflation cycle is a function of technology and the growth of global markets in goods and labour, is at most only considered in passing.

The depressing thing about this argument, is that the obvious conclusion would be that after a while we're just going to get washed way by another financial boom/bust cycle. Maybe not one whose epicentre is in the US and the UK to quite the same extent as the last one was, but that would be small comfort. It would be better to set global monetary policy with debt levels in mind, before the overall global debt/GDP level gets too much higher.