Market-Implied Population Growth Forecasts

Scott Sumner has written a huge amount about how an NGDP futures market would be an excellent thing. It would give the Fed something to target and allow us to easily get market-implied estimates of the effects of various events on the broader economy, including policy changes. And it looks like he might be successful.

The problem with suggests for various such markets in general is that there simply isn’t enough interest in trading them. The oil futures market exists because there are many private actors who wish to hedge their oil exposure; this is not the case for GDP. There just aren’t enough people interested in trading them.

What could change this is if the government (or someone else, I guess) decided to issue debt linked to GDP. Investors are always interested in pricing debt securities. You don’t need the possibility of hedging for the trading to create value – it can create value through time-value-of-money arbitrage, just like for equity and debt. This is after-all how we ended up with the TIPS market, and the extremely helpful market-implied inflation expectations (breakevens) it produces.

It seems like it should actually be quite attractive for the government to issue GDP-linked securities. Fixed-rate debt puts governments in a levered position, reliant on future growth to be able to fund themselves. When growth stalls, difficulty in repaying the debt is added to all their other problems. But GDP-linked debt would not have this problem; repayments would be highest when growth and thus ability to repay was highest, and lowest during times of recession.

None of this so far has been particularly new. Here is the new idea: population-linked bonds. The government would issue bonds whose coupons and face value were determined by the population of the country at time of repayment.

This would be attractive to the government, because it would help to match their revenues with expenses – periods of high population growth are likely to (ultimately) cause economic growth, and economic growth will encourage immigration. Conversely, low or negative population growth, which as Japan has seen ultimately lead to very poor economic growth, would be accompanied by low debt repayments, as would recessions which reduced immigration. If you wanted you could limit it to “working age population” or similar.

And this would be useful for forecasting as well – it would give us market-implied forecasts of population growth, which is a key assumption for planning many things, like the required capacity for infrastructure projects.

Population-linked-debt would also allow for the forecasting of some really unusual things. For example, suppose a terrible disease broke out in your country – the reaction of this debt would give you an estimate of the death toll. Or suppose two countries both had population-linked-debt. Looking at the co-movements, especially when changes were being announced to laws about immigration, would give you a prediction of future migration between the two countries. Pro-natalist policies could be judged against their impact on such bonds.

One disadvantage is it would reduce the government’s incentive to increase population growth. Similarly, GDP-linked bonds reduces the government’s incentive to encourage GDP-growth. Life is good, so this effect would be bad. This downside would scale with the volume of such population-linked-debt issued. The epistemic advantages of getting market-implied forecasts, on the other hand, only require a market large enough for liquidity. So we could avoid this by limiting the issuance volume. There are liquid markets for equities with market caps of less than a billion, so a few billion dollars worth of issuance a year should be plenty. On the other hand, this solution sounds dangerously like assuming a miracle – there’s little reason to think the government would choose to limit issuance to a level which didn’t distort their incentives.

2 thoughts on “Market-Implied Population Growth Forecasts

  1. Paul says:

    If I want to know X to make plans, then it seems like I have a hedging interest in X. So why can’t governments who need to know X directly finance speculation on it, i.e. run a prediction market or buy insurance at auction?

    If you issue debt tied to X, it seems like you must necessarily pay the same cost in terms of (potentially very slightly) lower interest rates. It seems like every dollar someone spends thinking about the problem must be a dollar you lose, one way or another (barring investor irrationality); if your debt isn’t cheap, why did they waste time thinking about it? If the amount of thinking they have to do is so small as to be negligible, then that seems to make other mechanisms work just as well. Similarly, adverse selection concerns seem identical in the two cases.

    So maybe there is some logistical reason to do this with debt, but setting aside frictions I think it only looks attractive because it successfully obscures the real cost. It seems like the main issue is really that buying and selling bonds is legal and understood.

    The case of NGDP is especially amusing on this perspective; as far as I know the story for NGDP targeting is that *everyone* has a hedging interest in GDP, and their inability to hedge is a major mechanism by which monetary policy operates. I understand that there are serious frictions, but I do often feel that directly addressing this issue would be simpler and probably more effective (especially since NGDP really isn’t what they want to hedge against…)


    • Thanks for your comment. You make a good point about having to pay the costs of research, one way or another.

      One advantage of debt is it is continually traded, so we can see how expected population growth changes over time. We can also look at how it changes in response to surprising news e.g. wars, new government policies, or employment data. It also seems like we could get this research ‘on the cheap’. Once a hedge fund has invested the resources in developing a model to predict population growth, which is necessary for auction or debt, it’s relatively cheap to continue to run it and trade it.

      I don’t really know enough about prediction markets to comment. I would worry about how to subsidize them in an incentives-efficient manner.

      It is a mystery why people don’t even link their contracts to inflation. They do in Chile.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s