I’m intrigued by inflation; it’s yet another output of a complicated, nonlinear system that plays an important role in the economy and that we don’t understand very well.

Baskets and burgers as a measure of inflation

One of the challenges of working with inflation is that it’s hard to measure across economies and through time. One common way to measure of inflation is to use the consumer price index (CPI). The CPI is constructed by tracking a basket of common items (food, gasoline, rent, etc.) between regions to provide an idea of how prices differ between them and how they develop over time. Inflation can be calculated as the change in the CPI between periods (CPI inflation). 

CPI inflation is however only a rough measure of inflation since it’s partially region- and time-dependent. For example, people living in different regions consider different items to be ‘common’ (the weekly needs of an older rural farmer will be different from those of the young socialite couple in the city), which makes it hard to compare like with like. Similarly, the needs of people living at different times will also be different; the items that we consider common today (TVs, air travel, convenience foods) were less common in the past (the weekly needs of a young suburban family in the 1950s are different from the needs of the same family today). The CPI can accommodate such differences to some extent by creating different baskets for rural and urban consumers, and by acknowledging that basket substitutions are inevitable as consumer needs change over time. 

Economic growth and technological change will however also affect how inflation is measured. Today, for example, we benefit from both the low-cost labour afforded by globalisation and low-cost technology as a result of technological innovation. As a result, many of the common items that are tracked by the CPI have actually declined in price. This trend is thought to be at least partially responsible for the low inflation we’ve been seeing in the US over the last 25 years or so.

Despite CPI inflation being low, this doesn’t rhyme well with a lot of people’s experiences of escalating living costs. For example, housing is more expensive today than it used to be. Education is also more expensive, as is childcare. Healthcare costs are spiralling out of control. If we define inflation as the loss of a currency’s purchasing power over time (as more monetary units are needed to acquire the same product or service), how can we reconcile these different observations; that living costs are going up while the purchasing power of the currency (as measured) isn’t going down particularly fast? The simplest answer to this question is that we’re not measuring inflation correctly, despite our best efforts. So, what can be done? 

Ideally, you’d like to track a representative cross-section of the economy to capture all the different inputs that go into setting a price (domestic and commercial rents, transportation and energy costs, commodity prices, salaries and living and employment costs, and so on)—and to do so better than the CPI basket can do: The CPI basket is flawed because its constitution changes over time. Conversely, if you were to track the same product, you would be able to create a better measure of how prices change (especially if the price of the product is determined by the costs of its inputs and production). This measure would then be possible to use as a gauge—a stable yardstick—to hold up to the economy and to see how prices differ and change relative to the gauge. 

With such considerations in mind The Economist magazine has created something they call The Big Mac Index; using the price of the McDonald’s Big Mac as a gauge to hold up to the economy. (Seen through this lens, a Big Mac should always be worth the same, and if the price of the Big Mac changes it’s because the costs of its production have changed.) There’s something beautifully simple about this approach (even if it’s of course going to be flawed in other ways).

The Big Mac as a gauge of purchasing power: How much wealth can a burger buy you?

Because of my interest in inflation and feeling that CPI inflation doesn’t capture it very well, I have been looking for a good gauge. Since The Economist’s Big Mac Index is such a gauge, I figured ‘why not’, and downloaded a spreadsheet with Big Mac prices in the US since 2000 from GitHub and started to play with the numbers. Right out of the gate it was obvious that the price of a Big Mac has gone up a lot: While a Big Mac cost $2.51 in 2000, it was $5.67 in early 2020! In other words, the price of the Big Mac has more than doubled in the last 20 years. Put another way, this means that while one dollar would have bought you 0.4 Big Macs in 2000, that same dollar will only buy you 0.18 Big Macs today. Such is the loss of purchasing power because of inflation.

Plotting this loss of purchasing power (number of Big Macs you get per dollar) offers an intriguing observation: The number of Big Macs that you get for a dollar follows a power-law-like relationship with a high degree of correlation (r² = 0.99). In other words, the price of a Big Mac can be predicted ahead of time. Therefore, if this correlation holds, the amount of dollars needed to get you a Big Mac seems to double every 20 years or so, so that the Big Mac that cost $5 in 2016 will cost $10 in 2035, and $20 in 2055. At the same time, the Big Mac itself shouldn’t change (a Big Mac is a Big Mac is a Big Mac). Instead, something is happening to the dollar—and happening faster than what is captured by the CPI. (While the price of a Big Mac correlates with US urban CPI, CPI inflation has been lower than the price-increase of the Big Mac: US urban CPI was 172 in 2000 and 257 in 2020; an increase of only 1.5×.) The Big Mac might therefore afford us a slightly different perspective on inflation.

(Now, Big Burger might have a spreadsheet in their HQs looking something like this, which they use for pricing their burgers. This can also be a measure of how inflation is eroding the purchasing power of the dollar. The truth is probably somewhere in-between. However, let’s assume that the Big Mac can tell us something about inflation, for now.)

If we use the Big Mac as a gauge (assuming that the value of a Big Mac remains stable over time), we can start expressing the cost of other assets in units of Big Macs. All of a sudden, we can now see how the price of other assets are changing not in dollars (which lose purchasing power over time), but in Big Macs (whose value we’re assuming will remain stable over time). Now, if we do this, we start making more interesting observations. For example, the number of Big Macs you need to buy the average US home (~70,000) has remained steady since the Great Financial Crisis. (The housing bubble can also be clearly seen as an increase in the number of Big Macs needed to buy the average US home; up to 100,000 Big Macs at its peak.) This suggests that if you were paid in Big Macs, the housing market would be as-welcoming in 2020 as it was in 2009 as the price hasn’t changed. Now, that’sinteresting: This doesn’t rhyme well with our impression of escalating housing costs.

Another thing that feels like it’s been rising in value is the stock market. However, if we plot the number of Big Macs needed to buy one unit of the S&P500 (which I like to use as a US stock market proxy), we’ll see that the number of Big Macs per S&P500 unit has remained relatively stable over time at ~400 Big Macs…! (The dot-com bubble, the dot-com burst, and the Great Financial Crisis are clearly visible.) The correlation between how many Big Macs you get per dollar and how many S&P500 units you get for the same dollar isn’t super-strong (but still meaningful) at r² ~ 0.7. This suggests that there is a (simple, absolute) relationship between the dollar-cost of a unit of the S&P500 and the dollar-cost of a Big Mac, so that if one goes up, the other will go up too. In other words, it’s possible that some of the increase in the dollar-price of a Big Mac is related to the price-increase of the stock market.

Two types of inflation and the burger as a lifebuoy

I’ve been wondering for some time now if the US economy isn’t seeing asset-price inflation (as opposed to consumer-price inflation) and if that’s not why CPI inflation seems to be so out of sync with people’s actual experience of life in the US economy. According to this interpretation, inflation isn’t happening in the cost of goods and services tracked by the CPI basket as much as it’s happening in pure assets (like stocks and real estate). 

The drivers of this kind of asset inflation would be many and complex. Many people have been crying foul of the Fed, for example, for having kept interest rates artificially low. This might have incentivised more and more money to go into the stock market rather than the economy proper (for reasons that will form the topic of another blog post). When more and more money is chasing the same asset (stock market units, say), the price of those assets will, however, understandably, go up. If more money is chasing fewer assets in the economy per se, it’s likely that we will see CPI inflation instead. (The 1970s oil crisis and resultant inflation would be a case-in-point.) Conversely, if more money is chasing fewer assets in the more abstract, asset-based economy (like real estate or the stock market), it is less likely that we will see CPI inflation. Instead, asset-price inflation should result. 

As the purchasing power of the currency is declining, the value of the goods and assets themselves will however not change (assuming a stable supply and no technological innovation). Thus, the value of a loaf of bread should always be worth the same, even as the price of the loaf might change over time. The same is true of assets (assuming stable supply and production costs), even as their role as investments complicate this relationship somewhat. (C.f. gold.)

In an economy seeing CPI inflation, the people with the lowest incomes will be the hardest hit as they need to spend more and more of their income on the same amount of goods. As the purchasing power of the currency is declining, the value of the goods and assets in the economy however doesn’t change. (More monetary units will just be needed to secure the same amount of goods.) This sort of inflation will burn from the bottom up: As the cost of goods increases, so will the price of assets, and incomes will be degraded unless they keep up with inflation. This type of inflation will affect people equally (even as people with higher incomes will have more of a buffer and feel the absolute effects of the inflation less than those with lower incomes, even as the relativeimpact is the same). 

In an economy seeing asset-price inflation, people with different asset-price exposures will however see their purchasing power develop differently: As the price of assets (but not consumable goods) is increasing, the people with the highest incomes will benefit as the asset-price inflation will cause their relative incomes to increase (as more of their wealth is invested in assets and therefore tied to their value). Therefore, as asset-prices go up, so will the wealth of people whose income is tied to the assets themselves. People without asset-dependent incomes will not see their income increase, and, as a result, even as their absolute income remains the same, their relative income will go down.

This is interesting, because this is pretty similar to what is currently happening in the US economy: Where people with low incomes are seeing their cost of living going up even as their incomes have stayed the same. Conversely, people with higher incomes (of which more is invested in assets and thus linked to asset values) have seen their income (and wealth) increase. As a result, their cost of living has remained more stable over time. Another way to see this would be to think of people with high incomes and wealth invested in assets as having a lifebuoy tied around their waist. As asset-prices increase because of asset-price inflation, the lifebuoy (the asset-dependent investment) will see them float on top of the asset-price inflation. (From their perspective, not much as changed.) For people without investments in assets (and thus no asset-dependent lifebuoys), the asset-price inflation will leave them in the same place—but worse off—as they started: They have no lifebuoy to keep them afloat. Altogether, this means that asset-price inflation will (in relative terms) make the rich richer and the poor poorer, which fits very well with what we’ve been seeing in the US economy.

Why, then, has the price of assets been increasing? Here, I actually think that the people who criticise the Fed have a point: By lowering interest rates artificially (below the level that they would be at in a free-market system), the Fed has incentivised money to move from savings (in, for example, bank accounts and bonds) to move into the stock market (which inflates the price of market assets). In addition, economic stimulus in the form of bond repurchases and QE has seen the amount of money in the asset-connected economy going up. In addition, I think growing pension assets might have an additional role to play here, too (partially by exacerbating the impact of the previous examples). If this is the case, the rich getting richer might be an unintended effect of past decades’ stimulus to encourage continued economic growth. Ironically, this stimulus might, therefore, have had exactly the opposite effect. 

So, what does all of this have to do with Big Macs? Not much. The Big Mac is just another way of expressing relationships between different goods and assets in the economy. You can reach many of these same conclusions by relying simply on the CPI, but I find the Big Mac to be a more intuitive tool. It also seems to capture more than the CPI does (even if it probably misses other things instead). As such, it makes a useful complement to CPI-based inflation metrics. However, expressed in this way, there are two obvious conclusions to be made: First, you want your income to be tied to assets (so your income is tied to asset-price inflation) or something that goes up alongside it (like the Big Mac), so you don’t lose purchasing power over time. (This loss of purchasing power might however be the price of the liquidity afforded by using a currency that is a derivative of an asset rather than the asset itself.) Second, the way to improve income inequality in the US will be to equalise the asset-base that people’s incomes are pegged to so that asset-price inflation won’t cause the high-income earners to float away from the low-income earners. How to do so constructively (and avoiding ironic side-effects) is the really hard but interesting question, and it’s one I don’t have a good answer to right now. (For example, having everyone own their own home turned out to not be such a great idea, as the housing bubble and the Great Financial crisis showed.)