What do a barrel of Belgian ale and a barrel of feta cheese have in common? How about a dental-extraction forceps and a Rolex watch?
Dying to find out???
Ok ok, I’ll give it away. It’s their levels of “pass-through”: That is the degree to which a depreciation of the dollar will translate into higher prices for you, me and your dentist. As it turns out, the prices we pay for beer and feta are less sensitive to changes in the dollar’s exchange rate (they have a low pass-through), while prices of forceps and watches are more so.
So what determines the level of pass-through? And who cares?
First question first. Since 2002, the dollar has declined 42 percent against the euro and 25 percent against a basket comprising the currencies of America’s main trading partners. Now, you would expect consumer prices to rise as a result—at the very least because imported products (from Belgian ale to dental forceps) should be more expensive.
So let’s see what happened: Looking at the consumer price index (or CPI), the cumulative increase over the same period has been 22 percent. Hmmm, that’s close to the dollar’s depreciation against the trade-weighted currency basket.
But don’t jump into conclusions: Much of US inflation in recent years has been largely driven by huge increases in the world prices of food and energy. As an example, the import prices of fuel and energy products have risen a staggering 530 percent over this period—way larger than the dollar’s depreciation. This means that if you isolate the impact of the dollar’s depreciation, you’ll find it to have been fairly small. In other words, you’ll find a low pass-through.
Indeed, studies looking at industrialized countries have shown that the exchange rate pass-through to prices is generally small, industry-dependent, and significantly less than one for one (meaning, if the exchange rate falls by, say, 10 percent, prices will rise by considerably less than that). Importantly, pass-though has been lower in the United States than in other countries. How come?
First of all, there are two stages of “pass-through”: The pass-through from the dollar’s depreciation to import prices; and the pass-through from the change in import prices to retail prices. Let’s take a look at each one:
From Belgium with love: Think of a Belgian trappist monk who aspires to elevate his obscure ale, Westvleteren, to a household name in the US. Apart from the fact that he might want to change the ale’s name into something more… pronounceable, the monk decides it would be a good idea if his export prices were not completely out of whack with those of his American competitors.
That was fine back in the good old days when the euro was trading one to one to the dollar: He could sell his six-pack for $8.99 in America and make roughly the same as he made in Brussels—9 euros per pack. But what does he do now that one euro is worth 1.55 dollars? Does he raise his prices to $13.95 to preserve his profit margins in euros? Or does he take a profit hit?
Turns out that most exporters choose to take a hit—they prefer to cut their margins and preserve their market share, rather than raise prices in line with the change in the exchange rate and lose out to competition. That’s what we call “pricing to market.” And that’s partly why the exchange rate pass-through to prices is fairly low, particularly in countries such as the United States, where markets are vast and consumers reliably compulsive.
A Babel affair: Another reason for low pass-through has to do with the globalization of production—the fact that to produce, say, a t-shirt these days, you start in a cotton plant in Texas, move to a factory in Shanghai, then jump to Hong Kong for some finishing touches before your “Economists Rock!” t-shirt travels back to your local Walmart store 1.
Why does this matter? It does because not all exchange rates move together at the same time. For example, while the Chinese renminbi has been appreciating (begrudgingly!) against the dollar in recent years, Hong Kong’s dollar has not moved (because it’s pegged to the US dollar). So import price of the t-shirt would rise only party, since only one stage of production is priced in a currency that moved against the dollar.
Port-to-door: How about the pass-through from import prices to the retail stores? Economists have estimated that storage costs, transport costs and wholesale and retail costs tend to account for between 30-50 percent of a product’s final price 2. This means that, to the extent that distribution costs remain fairly stable over time, they act as “pass-through absorbers:” They cushion the final price from changes in the exchange rate.
So pass-through has been low… But could it rise? And going back to my original question… Who cares? Well, the Fed does.
Taking it personally: You see, the Fed takes the level of pass-through a bit… personally. Here is why: Suppose the dollar depreciates because of somoething beyond the Fed’s control–say, oil prices shot up and the US trade deficit widened. If the Fed has a solid track record as a guardian against inflation, the Belgian monks will know the Fed won’t allow the dollar’s move to affect general prices and they will not raise the prices of their trappist brew.
But what if the dollar’s depreciation is perceived as the outcome of irresponsible Fed policies?!! Like… lowering interest rates too much to boost employment?! Everyone, from the foreign monks to the local ale producers, will expect broader prices to go up, and they will be more ready to pass on the dollar’s depreciation to the final prices.
In effect, the Fed sees a low level of pass-through as a signal of people’s trust in its capacity and willingness to fight inflation… (and vice versa)! So if there were evidence that pass-through has recently increased, the Fed might tilt closer towards raising interest rates in the near future. And that’s when I start to care. You should too, by the way, and you might also want to tell your dentist.
1 For a colorful account of the globalization of production see the “Travels of a t-shirt” by Pietra Rivoli
2 Jose Manuel Campa and Linda Goldberg have done interesting research in this regard. See here if you’re curious
Originally posted at Models & Agents and reproduced here with the author’s permission.