So earlier this week the Reserve Bank of Zimbabwe conducted it’s first foreign currency auction and it went well by any measure. As soon as it was over everybody wanted to see one key metric: the new official exchange rate between the Zimbabwean dollar and the USD. The curiosity was understandable, the Reserve Bank of Zimbabwe had stubbornly kept the rate fixed for the past three or so months. But I have news for you, that new official rate is not as useful as you think it is.

It has everything to do with how foreign currency is being auctioned on the formal market. Here is how it works.

The Dutch Auction system

First of all to take part you have to meet some stringent requirements set out by the Reserve Bank of Zimbabwe. These include technical requirements such as a minimum bid amount of $50 000 i.e. you are only allowed to buy a minimum of $50 000 and not a cent less otherwise the auction would become bogged down and difficult to manage. There are also requirements that ensure that the foreign currency you are buying is for a good cause and so on.

Once you have cleared those hurdles you can now take part in the auction. During each auction, the RBZ has a pool of foreign currency that it avails. That’s right there is just one (electronic) bag of foreign currency available. This week it was US $10 345 250.04. Companies from various sectors including:

  • Food and beverages
  • Services
  • Consumables
  • Medical

As well as other sectors took part in that auction.

During the auction each and every one of these companies taking part in the auction, potential buyers, enter their bids for the amount of foreign currency they want to purchase as well as the rate they are willing to pay. On Tuesday these rates started at $25.50 ZWL and went up to $100.00 ZWL. So, for example, Delta might send in a bid that says they want $100 000 and are willing to pay a rate of $100.00 for it.

Once all the bids are submitted, the allotted amount($10 million in this case) is then assigned to the bidders from the highest bids down, until all of the whole allotted batch of foreign currency is assigned. The Dutch Auction starts from the highest bid downwards. The final rate that each buyer pays really depends on the exact nature of the auction but in this case, it seems those who bid the highest amount got the foreign currency they asked for, then the next highest batch and so on.

So for example the auction starts at $100.00:

  • Several companies feel this is acceptable and gobble up $2 000 000 US in foreign currency
  • The system then reduces the rate to $80 and several companies gobble that up at the trough until they have had their fill and there is $5 000 000 left
  • The rate is reduced to $60 ZWL and so on and each time the rate is reduced more and more of the foreign currency is taken up
  • Until maybe there is $1000 left which is finally gobbled up at $25.50 ZWL

This is all hypothetical in terms of the amount but that is the gist of a Dutch Auction.

The stupidity of averages

From here a weighted average (mean) is then calculated and the RBZ got a rate of $57.36 ZWL as to 1 USD. The reality however is that different companies probably paid a different rate which can be far higher or far lower than this average rate. From the data we have bids ranged from $25.50 to $100.00. That is a massive gulf that is smoothed over by the average of $57.36. It means it’s possible one company paid an official rate that is $44 ZWL higher than the average. It is also possible another company paid as much as $36 below the average rate.

Anyone who has taken a basic statistics course will tell you why averages, especially means are stupid:

  • They hide extremes, in this case, $25.50 versus $100.00 are all hidden in a neat figure of $57.36
  • Often what you are not getting is not typical data with means you get things like 3.5 people. In this case, it’s doubtful anyone paid $57.36 for their forex. It would be surprising if anyone paid this exact rate.

That makes the rate useful but meaningless

The rate is useful as an indicator of foreign currency demand in the country and so on but do not expect companies to use it. It’s not their cost of doing business and therefore it would be ridiculous to use it as such. Businesses, by convention only consider the real costs they have incurred otherwise their records become a work of fiction.

ZERA’s mysterious calculations

That is why a lot of people were mystified by ZERA’s latest fuel calculations. Despite whatever justifications they may give, using a rate of $57 to compute the new price of fuel has little merit. Is this the actual cost of acquiring the foreign currency that was subsequently used to acquire the fuel? If not then what is the logical basis of using what is essentially an average rate? That’s the reason why the price of ethanol is not using the rate of $57 because everything else there is just a work of fiction sadly.

It is understandable what ZERA is trying to do, they want to make sure they reduce the government’s subsidy burden. However, when the subsidy on fuel foreign currency is eventually removed they can no longer make such prescriptive documents since the actual cost of acquiring the foreign currency will be different for each company.

Perhaps with time…

If the auctions keep happening it is very possible that the differences between the highest and lowest bids will start to shrink. However, that’s not always guaranteed. There is a foreign currency shortage and each bidding company wants to make sure they get as much foreign currency as possible. They will submit bids that are as high as possible while some companies are more constrained with regards to what bid they can submit.

Ultimately the solution is always going to be improving foreign currency reserves and reducing reliance on imports.