Nothing Can Hedge Against Inflation

We have already discussed how and why inflation is bad for you, investment-wise. All returns should only be considered after inflation has been factored in: it’s the real return that matters, not nominal.

 Although it’s the general level of inflation that is talked/tweeted/worried about, it’s actually specific inflation that’s the problem (I don’t care about petrol prices rising if I don’t drive). It’s best to try and counter this specific problem with a specific solution; real inflation protection is about hedging future consumption. Your specific future consumption.

The problem is that this is easier said than done. How do I hedge against the rent on my dingy 7th floor flat in South London rising exponentially? I could purchase some type of retail-focused, London-focused, “up-and-coming”-focused REIT but this doesn’t give me much protection against micro-area-specific demand-supply imbalances that inevitably lead to price movements in the upward direction.

So, alas, I find myself giving general inflation-protecting assets second consideration. I have already discussed why these assets may shield against the corrosive influence of inflation on returns but I thought I’d take a look at if they actually do; if they would have provided protection in the past, to what extent, and how precisely.

Let’s look at the data

My rigorous method of deciding which assets to assess involved recalling which assets I had seen quoted as being a “hedge against inflation”:

  • The stock market
  • Owning property
  • Oil
  • Gold
  • Cryptocurrency

Wouldn’t it be just lovely if one or more of these assets would increase each month with the general level of prices? Your expenses would tick up slightly each month but don’t worry – your assets have grown in value too.

Note: CPI refers to the Consumer Price Index and HPI refers to the House Price Index. Source: FRED, ONS.
Note: CPI refers to the Consumer Price Index and HPI refers to the House Price Index. Source: FRED, ONS.

As we can see here, house prices and the general level of prices seem to be moving in tandem directionally. General prices continue to rise in a slow, monotonous, inevitable fashion (in recent history) but house prices take you on a bit of a ride.

One could conclude that housing is therefore a reasonable hedge against inflation. But this appearance is largely a mirage, caused by the fact that both have simply been moving up and to the left since the start of our sample period. This is the problem with this type of analysis in general – any asset that has been increasing in price over the last X number of years will appear to be a hedge against inflation because prices have been rising essentially since we started recording prices.

When we start to look at the actual numbers we see that all these assets are not fit for purpose. The correlation between inflation and returns is slightly better than what you might find between inflation and noise:

HPIFTSE 100GoldOilBitcoinEthereumNoise
Data Start Year2005200119791987201420172005
Correlation0.15-0.070.020.160.170.110.08
% Changes Greater Than
Inflation Rate
60564953616345
% Changes Same
Direction As Inflation
59495256504749
Notes: These monthly changes are compared to the change in the CPI (Consumer Price Index) which is used a measure of inflation. HPI refers to the House Price Index. Oil, Bitcoin, and Ethereum prices were converted from USD to GBP. Noise were samples from U~(-0.05, 0.05) – I ran it about 5 times to get this outcome. Source: BoE, FRED, ONS, CoinDesk, Macrotrends.

As you can see most do not offer consistent protection on a monthly basis. 

WE NEED MORE DATA

“But Haydn,” I hear you cry “this analysis is unfair. Firstly – it would be unreasonable to expect an asset to track inflation on a monthly basis. Secondly – we don’t care; we just want something that will hedge against inflation in the long-run.”

Fine – let’s look at annual data and extend the timeline.

Note: General Prices refers to RPI and House Prices refers to a House Price Index supplied by Nationwide. Source: BoE, Nationwide.
Note: General Prices refers to RPI and House Prices refers to a House Price Index supplied by Nationwide. Source: BoE, Nationwide.

Again, it looks like house prices might be a good hedge. Again, the numbers say otherwise:

House PricesShare PricesGoldOilNoise
Data Start Year19531800197919871800
Correlation0.170.11-0.020.130.13
% Greater Than
Inflation Rate
6156545042
% Same Direction
Change As Inflation
8759625350
Notes: These annual changes are compared to the change in the RPI (Retail Price Index) which is used a measure of inflation. Oil prices were converted from USD to GBP. Noise were samples from U~(-0.1, 0.1) – I ran it about 5 times to get this outcome. Source: BoE, FRED, Nationwide, Macrotrends.

Annual still too short? Let’s look at 5-year buckets:

Note: General Prices refers to RPI and Oil refers to Crude Oil Prices: Brent – Europe, converted from USD to GBP. Source: BoE, FRED, Macrotrends.
Note: General Prices refers to RPI and Gold refers to Gold Price as recorded by the BoE. Source: BoE.

Why not 20…or 50?

Note: General Prices refers to RPI and House Prices refers to a House Price Index supplied by Nationwide. Source: BoE, Nationwide.
Note: General Prices refers to RPI and Share Prices refers to the price of a collection of shares that characterised the market for shares at the time of record, defined by the BoE. Source: BoE, Nationwide.
Uncertainty: the inevitable destination

The problem with this type of analysis is that lots of things determine the price of oil, house prices, Ethereum, etc. and lots of things also determine the rate of inflation. So any type of standard statistical analysis just isn’t going to cut the mustard in terms of discovering a relationship between these types of variables.

Low correlation doesn’t mean that we can rule out the presence of causality. But that’s ok – we don’t care about causality. We just want something that hedges against inflation – some price of some thing that moves in the same direction as the price of things that we buy. None of these assets, on a monthly or annual frequency, accomplish this on a consistent basis. Not to any degree we would like, anyway.

Whether this is true over longer periods of time is more unclear. Extending our data back hundreds of years might reveal that assets we thought were a “sure thing” in terms of countering inflation may not actually be that certain. And of course, we run into the age-old problem of induction mentioned too many times on this blog.