Making sense of the Weighted Average Exchange Rate


Earlier this week, the Central bank reintroduced the interbank foreign currency trading platform but with modifications. When the interbank was first introduced in February of 2019, it marked Zimbabwe’s shift from a fixed exchange system of 1:1 between the USD and RTGS (local currency) balances. Essentially, maintaining a fixed exchange rate meant that RBZ had to clear all outstanding obligations at that respective exchange rate and these included outstanding loan payments, dividends, procurement of inputs such as fuel and electricity and portfolio divestments by foreigners. However this money had to be generated through exports, so more significantly you need a relatively higher export level (source of forex) compared to imports, to sustain the model. Unfortunately Zimbabwe’s trade balance has been adverse in each year since 2000. More so, fixing the exchange rate typically discourages export production, as retentions from their sales due to government are compensated at sub market rates.

Without digressing much from the subject of this piece; the fixed exchange rate system which was in place between 2015 and 2018 failed because it discouraged production and free market forces, the balance of payments was worsening as import bill grew ahead of exports receipts and more importantly, a sustained budgetary imbalance led to proliferation of local balances created through government’s open market operations (Treasury Bills) as authorities sought liquidity from the local market to satisfy its budgetary excesses. Fast forward to 2020, there have been attempts to allow for market forces to determine the fair value of the Zimdollar against other currencies. Crude data from 2019 shows that average daily trades on the interbank for the year were in the range of US$2.2 million a session. This figure compares to average daily demand of at least US$15 million and potential supply of US$12.5 million a session. The demand and supply levels are based on historical trade data. Using the inferred data, it is clear that the level of variance in actual trades to both potential demand and supply levels is very wide, which demonstrates that the market was somehow not working.

Indeed it was not working as the RBZ initially did not let market forces determine the rate. The Bank put caps on the daily exchange rate, which means it dictated that the rate could only fall as much in a single session, which essentially is a control measure and digression from market forces. Gradually it gave rise to off-market trades in twinning arrangements, which were executed at more favourable rates between importers and sellers with the help of banks. A study done by Equity Axis showed that periods of excessive RBZ controls were accompanied by very conservative exchange rates or mild weakness in the ZWL$ but this was matched with the opportunity cost of volumes. Trades on the interbank spontaneously dwindled in periods of excessive control. The opposite was equally true. Looked at from a relative perspective, in periods of excessive RBZ controls, the parallel market rates for the local currency fell at a much quicker pace than periods of less control. This shows that developments on the interbank have an effect on the parallel rate.

Against this background, should we expect a different result from the interbank market than that achieved in the past. From the above it is clear that effectiveness of the market it is largely dependent on RBZ, its influence or lack thereof. Let us now look at the modified interbank market and break down its mechanics. The refined trading system comes along a Reuters platform, which is basically a more transparent auction system bringing together all primary dealers and the RBZ. Primary dealers are supposed to place bids and offers on behalf of their clients every week on Tuesday morning, on to the system. Through the auction market, bids are matched with offers. For example someone may be looking for US$1million at 70, the other US$5 million at 50, these bids can only be matched or satisfied if on the other side there are buyers willing to dispose at those same price levels, similar to stock market trading.

Assuming the above are the overall bids on a Tuesday, irrespective of what happens on the supply side a rate can be calculated taking from the RBZ’s new model. The weighted average exchange rate is calculated using the weight of bids, the exchange rate demanded by a buyer requiring a bigger parcel of forex carries more weight. So a bid of US$5 million @50 carries more weight compared to a bid of US$1 million at 70. The weighted average exchange rate will therefore tilt more towards the 50 bid than the 70. In this hypothetical example the weighted average exchange rate will come in at 53%. According to the new model this average weighted exchange rate becomes the focal rate for the week. All trades conducted between Wednesday and the coming Monday are all guided by the rate set on the preceding Tuesday. Trades can only hover up or down by circa 2.5% which is minimal variance to the focal average weighted weekly rate.

From an analysis perspective, there are quite profound observations. First, the standard for weekly auctions is not a global standard. Typically markets are liberalised and the rate is set on a real time basis. Central banks maintain forex and gold buffers as contingent in case a country’s currency becomes volatile. These reserves are sufficient cover to defend a currency once it’s under siege. For instance when there is a rush to buy forex and the local unit is tumbling, Central banks inject more forex buying the local unit, which in turn reduces the volume of local unit in the system. Effectively this may help rebalance the market through pure market forces and not physical force. Zimbabwe readily does not have any forex or gold reserves to cushion currency volatility. Reserves of at least 3 months import cover are recommended as a way of reducing currency volatility. Zimbabwe would therefore require about US$1.2 billion in forex reserves to cushion the currency, this is an equivalent of a year’s gold exports receipts.

Weekly auctions therefore act as circuit breakers dictating the pace of weekly depreciation or appreciation in currency. The Zimdollar has so far moved in one direction which is southward. A bigger challenge here is that the rate set on Tuesday is a fair representation of the value of local currency. Unlike the stock market where such a model fairly represent the true state of market, the interbank weekly rate has technicalities that influences and may misrepresent the fair value. As a player, the Central Ban has a right to participate in buying and selling of forex on the IBM. In the past, the Bank promised to supply 50% of all its retentions to the Interbank. So there is a scenario where the Bank may procure forex from exporters at sub optimal conservative rates, depriving the exporter and yet channel same at weak prices to the Interbank as a way of keeping the rate down. Once bids as low as 25 are satisfied, successive auctions will still see very low bids with High volumes, which again drives the weighted average rate lower.

This scenario will lead to a rate that is not truly reflective of market forces for an entire week. Interestingly the fact that the auction is held once a week mean RBZ can play this card longer. It has a whole week to spare its arsenal and may strategically play the market only on the day of setting the rate. How will it achieve this. Since all other days between Wednesday and the following Tuesday will simply be using a rate already set, it means the Bank can retreat in-between and come back to defend the currency on Tuesday. Maybe these are mischievous thoughts but we have seen a lot as a country, all possible scenarios need to be analysed. This is clearly one of the shortcomings of the interbank market with weekly auctions as circuit breakers. The Bank can manipulate the rate for an even very long time but as we know, as the variance to parallel market widen, sellers are discouraged and the Interbank will stop working. This scenario may also play out on the buy side on possible collusion.

One major drawback of the average weighted exchange rate is that it can be set even without supply, partial or total, by merely using bids and volumes of same. So to correctly analyse the rate, it is important to refer to supply level. For example this Tuesday, over $10 million was supplied and the bids ranged from 25 to 100. Demand at about $11 million was almost completely satisfied. If the supply levels vary widely to demand, it may imply that the market is in disequilibrium and to restore same the price of ZWL will have to fall. But again this was only the first week and to get a better picture of the market we need more time. You can imagine exporters unutilised balances had been spared from liquidation during lockdown and this had a consequence of accumulation of balances, winding these down may temporarily spur supply among other issues. History has shown that market forces will always ultimately prevail. It is however plausible that as an economy we are at least attempting to move in the right direction.


Please enter your comment!
Please enter your name here