The Mexican peso trades at a 50% discount on the streets of Peru and I believe this is somehow related to Peru being the top source of counterfeit U.S. dollars in the world.
That’s the basic thesis.
What is about to come next is a series of observations, some postulations, an educated guess or two, and a some rough idea of where to go next.
Contained nowhere in here is actual proof, in the Euclidean sense of the concept. But I’ll defend that by saying that what this brain dump is about is so off the radar that the only possible proof to have is to be an actual perpetrator. So, not being a perpetrator, what I can spell out here is my thought process, and I wholeheartedly welcome comment from anyone who can fill in any of the gaps I’ve left, or better yet, prove me wrong.
With that out of the way, let’s get to the data.
This whole idea began with a simple observation: upon landing in Lima’s airport, the first currency changers available—the ones you see before even clearing customs—offer to sell Peruvian soles in exchange for travelers’ American and Canadian dollars, Japanese yen, British pounds, euros, Swiss francs and Mexican pesos, quoted in terms of how many soles you can expect to receive per unit of your foreign currency. Considering that the first currency changers available upon landing at any airport anywhere in the world will never give the best exchange rate, the rates in Lima’s airport do not depart from current market rates as much as they could.
The single exception is the Mexican peso.
The number of Peruvian soles one can expect to receive per Mexican peso in the Lima airport is quoted as “0.1”. What does this mean?
To equate this to something understandable, we can use the old cross-canceling technique of multiplying fractions with different units. Here’s what happens when we do that:
1 Mexican peso 2.5 Peruvian soles 25 Mexican pesos
——————– X ———————- = ———————-
0.1 Peruvian soles 1 US dollar 1 US dollar
Yes, you’re reading that correctly: that’s 25 Mexican pesos per dollar. Look up the going market rate for Mexican pesos, and you’ll see them trading, as of this writing, somewhere in the neighborhood of 12.3 to the dollar.
So, anyone landing in Lima wanting to change Mexican pesos for Peruvian soles at the airport will do so at an implied devaluation of the peso on the order of 50%.
Put another way: you’re going to spend twice as many Mexican pesos as you should to acquire Peruvian soles.
Granted, this is in the airport. Clearly the next step was to see if any casa de cambio in central Lima takes Mexican pesos. The answer to that question is yes, as should be expected of any fully convertible currency in world markets. The rate I was quoted at the casa de cambio around the corner from my hotel was 20 Mexican pesos per dollar. This is better than 25, but not by much; the implied devaluation here is still nearly approaching 40%.
Now, one could argue there is fault in my sample size, that the data only came from the Miraflores neighborhood, which skews it this way or that way…in short, there are always more data points to gather. But for these purposes, for what I’m trying to establish, I do not believe more evidence beyond the airport and the wealthiest neighborhood in Lima is going to materially alter the findings.
Meanwhile, changing US dollars on the street in Lima can be done for very little, if any divergence from the market rate. The single data point I have here—drawn from the casa de cambio that was willing to take 20 Mexican pesos per US dollar—was receiving 258 Peruvian soles for $100 on April 23, which, according to oanda, was exactly the market rate that day.
To further demonstrate the relative demand for US dollars in Peru, my hotel bill for a recent three-night stay there was quoted as payable with 1,386 soles or US$462, which implies a USDPEN rate of 3.0. Given the market rate of 2.58 Peruvian soles per dollar, this means the hotel was willing to devalue its own currency by 14%.
I don’t know if the hotel would have been willing to take Mexican pesos, but that’s unnecessary for a preliminary conclusion here. The 50% devaluation penalty for changing Mexican pesos in Peru is more than just statistically significant. It’s basically the Peruvian foreign exchange establishment saying, “You’re either stupid or a drug dealer.”
Suppose it is. An alternative conclusion here could be that the more convertible a country’s currency is on an international scale, the less it will charge you when you try changing more “exotic” currencies for the local money.
On that front, I’ve since checked so-called “street” FX rates for Mexican pesos in New York, London, Toronto and São Paulo. In every one of those markets, the implied USDMXN rate ranges from 13.0 to 14.5, with the 13.0 rate coming unexpectedly from Sao Paulo. That São Paulo’s rate for pesos is actually better than New York, Toronto or London kind of goes against any theory that greater convertibility reduces the peso spread.
Where this goes next:
It’s probably worth checking Mexican peso rates in Bogotá as another frame of reference, not just for another Latin American country, but another transit point of known narco-related money laundering in the hemisphere. I have no idea what to expect here but am looking into it.
Drunkeynesian, who aided me with the São Paulo data, additionally offers this comment:
“I like your thesis, and I would add to it a premium for the difficulty to convert Mexican pesos in another “liquid” currency. In a country where there are many Mexicans (or people traveling to Mexico), rates should be better, since if you’re a FX broker and you get some Mexican pesos you won’t have to wait long until you make some profit. If this is valid, this premium should grow with the local interest rate (to compensate for the opportunity cost).”
To be continued.