Zubetune

Equity duration around the world
Some time ago, I wrote about the research of Jules van Binsbergen who showed that you can replicate the performance of the S&P 500 with a series of Treasury strips that pay the amount of the expected dividend at maturity. In essence, you replace uncertain future dividends with the certain payback of zero-coupon Treasuries and end up with less risk and the same or even higher return. And with lower volatility to boot. This research creates all kinds of problems from a theoretical standpoint because it purports to show that equity markets do not compensate investors for the risk of future dividend cuts or increases. And more importantly, it shows that stock market returns are essentially compensation for changing interest rates and not for any other risks. All that seems to matter for aggregate stock markets is where interest rates and inflation are going.

Back then I said we need to understand this result better, but while I have no answer except that as an investor, I am using expected changes in interest rates and inflation as the main, though not the only driver of my market outlook for the coming years, we now have results from three different international markets that show the same lack of equity market outperformance over a series of government bond designed to replicate expected dividend payouts.

In a follow-up paper van Binsbergen tested his methodology for the FTSE All-Share, the EuroStoxx, and the Nikkei 225 and came to the same conclusion as for the S&P 500. You can replicate the performance of the stock market at lower volatility with a series of government bonds. So, the problem of the lack of compensation for risk in equity markets seems more widespread than just the US market.

But also, van Binsbergen’s analysis allows us to calculate the sensitivity of different stock markets to local interest rates, i.e. the duration of the local stock market. The chart below shows the different durations of the stock markets. Markets with longer duration like the S&P 500 should do better in a world of declining and low interest rates. Markets with shorter duration like the FTSE and the EuroStoxx should do better in an environment of rising interest rates. This is simply a reflection of the dividend cash flows in different markets and the interest rate sensitivity of the stocks in the different markets.
As for this year, there seems almost universal agreement amongst investors that this will be a year of rising interest rates. And that means that as an equity investor, you might want to shorten the duration of your equity portfolio to reduce the negative impact of rising rates on the portfolio. The simplest way to do that is to sell US stocks and buy UK stocks.

But that is a problem in 2022 since it becomes increasingly clear that we are going to leave Covid restrictions behind and are returning to a more normal lifestyle that involves international travel and going out again. So right at the time when we are likely to start spending more on these goods and services again, the weights are reduced, thus creating a somewhat distorted picture of inflation going forward and – I guess – a somewhat lower headline inflation than we really experience.

Because of the different methodology in the United States that is not a problem there, and normally, the changes in weights are so small that they are not noticeable. But in many countries around the world the pandemic will lead to significant shifts in inflation baskets and make inflation number less representative and less comparable to the US numbers.

Why inflation data will be even less representative of your cost of living this year

Forgive me for writing about some really arcane area of macroeconomic statistics, but the pandemic is creating all kinds of weird effects in the macro world. For one, because of the massive disturbance from the pandemic to the global economy, macro forecasts are currently even less reliable than they normally are (which is saying a lot). But because of the way some macroeconomic aggregates are calculated, they are currently becoming less representative of what is really happening in the economy than before.

Earlier this year, the Office for National Statistics in the UK and the statistical offices across the EU published the new weights with which products and services are entering the basket to calculate consumer price inflation. Unlike in the United States, where weights of products and services are adjusted a little bit each month, in Europe and the UK, the weights are adjusted once a year based on the consumption patterns of households in the previous year or so and then left unchanged for 12 months.

But thanks to the pandemic, our consumption patterns have changed significantly over the last two years. Relative to other items, we spent less on transport (both for commuting and for pleasure). Less on restaurants and less on recreation and entertainment because – well – we couldn’t go out most of the time. And as we spent less on these items, the relative weight of other sectors like housing, food and clothing increased. The chart below shows the shift in weights in some of the sectors impacted by Covid in the UK that are put in place at the start of 2022.
If I add up the weights of sectors that have been adversely affected by Covid and the weights of sectors that have not been affected you end up with a pretty dramatic shift Away from Covid sensitive goods and services.

Change in sector weights of UK goods and services affected by Covid and those that weren’t
But that is a problem in 2022 since it becomes increasingly clear that we are going to leave Covid restrictions behind and are returning to a more normal lifestyle that involves international travel and going out again. So right at the time when we are likely to start spending more on these goods and services again, the weights are reduced, thus creating a somewhat distorted picture of inflation going forward and – I guess – a somewhat lower headline inflation than we really experience.

Because of the different methodology in the United States that is not a problem there, and normally, the changes in weights are so small that they are not noticeable. But in many countries around the world the pandemic will lead to significant shifts in inflation baskets and make inflation number less representative and less comparable to the US numbers.

2 plus 20 really doesn’t work for clients
There is an endless debate about the fees fund managers charge their clients and how much and what incentive structure is fair. In this regard, I am glad I found a summary of all the experimental research that has been done over the last decade in that area.

Typically, there are three different fee structures that are commonly used in practice:

A fixed fee, which can be either a fixed amount or a share of the assets under managements but it is largely independent of the investment success of the fund manager

An aligned fee, where the fund manager earns a fixed fee and invests his or her own money alongside the clients’ money, thus participating in losses should they occur

A convex fee, where the fund manager earns a fixed fee plus a share of the returns, but does not participate in the downside should returns become negative.

The beauty of lab experiments is that one can test how fund managers invest their clients’ money in each of these scenarios and if they take on more or less risk than under different fee structures.

The good news here is that fund managers care about their clients. They do invest similar to the way they invest their own money when working under a flat fee. If they are held accountable by clients for their past performance (e.g. through personal annual reviews) they tend to reduce risk taking if the client can review the performance and identify underperforming managers or complain to them. Notably, risk taking by fund managers is largely the same whether they operate under a flat fee or have an aligned incentive scheme where they invest their own money alongside clients’ money. This is essentially a reflection that fund managers are human and humans are social beings who crave the acceptance and approval of their fellow beings. And the more fund managers meet their clients, the more they get that approval (or not, if they lose money). This feedback from clients through personal approval seems to be as strong a force as investing personal wealth in their fund.

Where things really go wrong is once convex incentive structure come into play. When the fund manager shares in the upside but not in the downside, there is a strong incentive to take more risk and indeed, compared to a flat fee, fund managers operating under a convex fee structure take on substantially more risk. In one experiment, fund managers received a 5% performance fee on all the profits their clients had but faced no penalties if their clients lost money. The reaction of fund managers compared to a flat fee was to double their investments in risky assets.

And in lab experiments you can do what you can’t do in real life. So in a follow up experiment, the fund managers were rewarded with a bonus of 50% of their clients’ money if they made a profit. This created a negative expected return for their clients if the fund managers invested in risky assets, but guess what, the fund managers didn’t care. They invested the same amount in risky assets as under a convex fee structure even though they knew their clients would lose money.

There is a reason why I stay away from money managers who charge 2 plus 20 or similar convex fees.

Is this why cryptocurrencies are better than gold?

In October, I wrote a post about the fiasco that was Amazon’s New World game. They accidentally created a world where the supply of currency was limited and could not be increased. The result was a massive deflation and a collapse of the economy in that virtual world. I compared this scenario to what would happen if we return to a fixed gold standard or adopted a cryptocurrency like Bitcoin with its limited supply as standard.

As with every post I write about cryptocurrencies I got lots of responses from crypto advocates (and in this case also from gold advocates). The basic argument of gold advocates is that gold supply increases by a few percentage points each year from gold that is freshly mined. My counterargument is that this is true, but if you want to fight the collapse of an economy, you have to be able to temporarily increase the supply of money dramatically in order to prevent a run on banks or hoarding of good money (see my remarks on Gresham’s Law in the previous article). And economic crises don’t care if gold mines can produce enough output or not. So, a gold standard will inevitably lead to a deflationary economic collapse. This is why central banks had to abandon the gold standard in the Great Depression and why the financial crisis of 2008 would have been as bad as the Great Depression had we still been on the gold standard then.

But the crypto advocates had a different argument and one that is seemingly better. They argued that cryptocurrencies can adjust their supply with relative ease to react to changing demand. This is not true for Bitcoin with its stable demand, but it is true and actually happened for Ether. On 5 August 2021, the Ethereum Improvement Proposal 1559 (EIP-1559) was put into action. Since then, every new transaction that is included in the Ethereum blocks levies a base fee that is eventually destroyed. This means that the supply of new Ethereum tokens is curtailed by the base fee, thus creating a reduction in supply growth. This was done to prevent oversupply of Ethereum since this digital currency has no upper limit to how many Ether can be mined. But the same mechanism can be used in reverse to create additional supply should there be a need for it. Essentially, with a small coding change, the supply of Ethereum can be adjusted upwards or downwards at a moment’s notice.

This is great and should prevent a New World style deflationary collapse.

But assume we are using Ethereum as the standard currency. Who is going to decide when and by how much to increase or reduce the supply growth of Ethereum? The current methodology of democratic decision making about EIP is far too slow and too cumbersome to work in an emergency. Besides, assume that the UK adopts Ethereum as its currency. If the people in the UK decide that they need a higher supply of Ethereum but a majority of Ethereum holders in the world (most of whom will not be British or even live in the UK) decide that this is not a good idea, these foreigners can dictate monetary policy that applies to the UK. A democratic process run by owners of Ethereum worldwide clearly doesn’t work on a national level.

Thus, what one needs is a cryptocurrency that is controlled by the British and able to make fast decisions if there is an economic emergency. In order to make a quick decision one needs a group of experts that is competent enough to make the right decisions in the best interest of the people and that can act fast enough to implement changes in the supply of the cryptocurrency. And because they need to be able to act in the best long-term interest of people instead of every whim and populist call for more money, they should not be elected representatives of the people. We see in so many countries what happens if you elect populists.

But wait, don’t we have such a group of experts that are able to make quick decisions about the supply of currencies without being exposed to populist tendencies? Yes, we do. They are called central bankers. It is the reason why we have central banks in the first place because for centuries we have tried other ways of managing the money supply and they all failed. And this is why I am supportive of central bank digital currencies but not the dreams of cryptocurrency optimists who think that a decentralised currency will somehow be able to better for our economic wellbeing. I have not seen any solution to the problem of dealing with economic emergencies like the financial crisis, the Great Depression or the New World deflation crash coming out of the cryptocurrency world that is even remotely realistic. And I doubt there is one.