Is Minister Kaingu trying to tell us that government is broke?

This week, the Minister of Higher Education Dr Michael Kaingu announced that government would scrap off the payment of meal allowances (popularly known as “BC”) to students at the University of Zambia (UNZA) and the Copperbelt University (CBU). According to him, the government had decided to withdraw the allowance because it “has been abused and that it has over the years been a source of unrest at the two universities”. Meal allowances have been an institution at the two universities and I was lucky to get “BC” during the time I studied at UNZA. Life would have been unimaginably difficult without BC, coming as it did at a time when there was great financial uncertainty for my family. That's my wife's story too. And my sisters' story and countless others who have benefitted from BC over the years and gone on to join the ranks of tax paying citizens.

My point here is not to debate the Minister’s reasons for scrapping off the allowance, even though he doesn’t provide any evidence for his assertion. For instance, is he saying that the payment of meal allowance leads to unrest (for example, cash flash students go get drunk and then throw stones at cars)? Or is he saying that students become unruly when the meal allowance is not paid on time? The conclusions one can draw are radically different depending on what’s causing what. Nor do I want to debate the legality of the decision he’s taken. Others have done so. Nor do I want to be roped into a debate about the unfairness of having few students obtain meal allowances when students at other colleges and universities aren’t doing so. Of course it's unfair that only a minority of students in the country get the allowance. But scrapping it off for everybody is a terrible type of equality - it meets the fairness principle but most people, when pushed, would see that this is undesirable. Instead we should achieve equality by moving in the other direction: making sure that every student gets the allowance. But that’s a post for another day.     

What I want to point out here is that the government’s decision says something about the country's fiscal situation. The decision is controversial if not unpopular (have a look at the commentary on social media). So why would a government announce what appears to be an unpopular decision a few months before a general election? Election years are for dishing out largesse not scrapping it off.

The only answer I can think of is that the government’s finances are currently stretched. Have a look at the figure below that tracks annual interest payments on the country’s outstanding stock of domestic and external debt from 2012 to 2016.  

Source: Budget Speeches

The interest payments are in billions of Kwacha (the rebased currency).* Notice that in 2012, total debt service payments (both domestic and external) were budgeted at K3.1bn. By 2016, the amount budgeted for external debt (K3.6bn) and domestic debt (K3.5bn) are each, on their own, greater than the total budgeted for 2012! That is, total debt service payments between 2012 and 2016 have grown by an incredible 120% (with the external portion growing at 150% and the domestic portion growing at 105%). Also notice that 2016 is the first year where external debt service payments are greater than domestic debt payments – itself a reflection of the growing importance of external debt in our total debt stock. Lastly, notice the incredible steep rise in external debt service payments between 2015 (K2.4bn) and 2016 (K3.6bn). That is, an increase of K1.2bn in a single year! Wow. Even more daunting if you remember that external debt service payments are made in foreign currency.  

Now according to the 2016 budget, the total budgeted for “Student Loans and Bursaries” is K310million (this is a total figure that includes the meal allowance). It is far less complicated to cut funding to this or that social service than to miss an interest payment when a government is debt stressed (there’s actually academic evidence showing this). Missing an interest payment screams loud and clear that you are broke. You can't spin this any other way. On the other hand, you can spin cutting back social spending by simply saying that it was being wasted or it encouraged unruly behaviour (after all, who hasn't heard of the entitled car-smashing students). 

The money originally meant for student loans and bursaries doesn’t, however, come close to plugging the fiscal hole. Implication: Times ahead are gonna be tough. 

* The numbers presented here are in nominal terms. The distinction, in my opinion, between real and nominal doesn't matter so much here because inflation has been quite low over the period covered - the uptick in inflation only happened towards the end of last year after the budget for the current cycle had already been announced. 

Should BoZ be counting Eurobond proceeds as part of forex reserves?

I have noticed that the Bank of Zambia (BoZ) has been reporting the proceeds of Eurobond issuances as part of the country's stock of foreign exchange reserves. To illustrate, I report in Figure 1 the end-of-quarter forex reserves as reported by BoZ in their Fortnightly Statistics Bulletins. The data runs from the first quarter of 2012 (12q1) to the third quarter of 2015 (15q3).*


            Figure 1: Foreign Exchange Reserves, 2012 q1 to 2015 q3
 Sources: Bank of Zambia and Ministry of Finance


As is clear from Figure 1, build-ups in forex reserves have coincided with successful issuances of Eurobonds. For example, in the third quarter of 2012 (12q3), the country issued its first ever Eurobond valued at US$750mn. Coincidentally, in the third quarter of 2012, our stock of forex reserves increased from about US$2.5bn at the end of 12q2 to about US$3.3bn at the end of 12q3. Thereafter, there was a steep decline in forex reserves probably as the money was released for infrastructure projects and/or utilized to intervene in the forex market (I have covered the last point before).

The stock of reserves was largely flat in 2013. Thereafter, there is a steep rise in 14q2 (second quarter of 2014) when reserves rise from about US$2.7bn at the end of 14q1 to about US$3.4bn at the end of 14q2. Coincidentally, a second Eurobond of value US$1bn was issued in 14q2. The reserves are immediately drawn down for, perhaps, the same purposes reported in the previous paragraph.

Reserves are largely flat at around US$2.7bn from 15q1 to 15q2. There is an immediate steep rise in 15q3 when reserves rise again from about US$2.7bn at the end of 15q2 to about US$3.6bn at the end of 15q3. 15q3 coincides with the issuance of a third Eurobond worth US$1.25bn.

It appears then that the Central Bank is including the proceeds of Eurobonds in reporting the country's international reserve position. This point was sort of confirmed when the Minister of Finance, in his Budget Address to the National Assembly last Friday, had this to say:

[A]s at end-September 2015, Zambia’s gross international reserves stood at US $3.6 billion compared to US $3.1 billion recorded at end-December 2014, representing 4.4 months of import cover. This improvement was partly on account of inflows from the recently secured sovereign bond.

Should the Bank of Zambia be including Eurobond proceeds in reporting our forex reserves? This does not make much sense to me at a conceptual level given that forex reserves should really be thought of as assets. In other words, in building up reserves, the Central Bank should go into the forex market and *buy* and thereafter *own* the reserves. By including Eurobonds in their reporting, it seems to me that the Central Bank is counting what really is a liability as an asset. They are bonds after all, right?

Unless BoZ has been purchasing the Eurobond proceeds from Central Government, using Kwachas immediately when the money comes in from abroad. But even then, doing so would imply that these are really temporary assets since the primary reason for issuing the Eurobonds was so that we could use them for infrastructure projects. In other words, Central Government would still want to have the US dollars back at some point to use them for their intended purpose. That is, BoZ  would still be in the position of treating a liability as an asset.

The good folks at BoZ are way smarter than I am. There must be something that I am missing in my story. Can someone help me out?  

* The latest Fortnightly Statistics Bulletin from BoZ does not contain data on forex reserves for 15q3. The number for 15q3 reported in the Figure 1 is taken from the 2016 Budget Address by the Minister of Finance.

Some exchange rate charts and observations for Zambia

Zambia's currency, the Kwacha, has taken a bit of a beating these last couple of days. According to the currency converter Oanda.com, the currency opened last week at around K8.59 to the US dollar and closed the week at K9.54 to the dollar, a depreciation rate of around 11% in a single week! Oanda's interbank rate as at yesterday was K9.95 (almost K10) to the dollar.

Some commenters following the local currency's gyrations have made the point that the Kwacha's performance this year is not exclusively a Zambian thing. Other African currencies are also feeling the pressure of China's possible slowdown and/or monetary policy developments in the US.  So I decided to compare the Kwacha's performance against the performance of a selected number of African currencies - chosen because they are usually invoked by many as comparison currencies for the Kwacha. I have also included the performance of the Chilean Peso given that Chile, just like Zambia, is vulnerable to a slow-down in copper demand. (Readers are encouraged to do this comparison exercise for as many currencies as possible). The results of this exercise for the period 31st January 2015 to September 4th 2015 are in Figure 1 below.


                                           Figure 1

                                          Source: Oanda.com


What is clear from Figure 1 is that the Kwacha has depreciated the most when compared to the Naira, Chilean Peso, Kenyan Shilling, Ghanaian Cedi, Rand and Ugandan Shilling. The Kwacha has depreciated by about 50% over the period January 2015 to early September 2015. Only the Ugandan Shilling comes close with a depreciation rate of around 30% over this period.

What is clear from Figure 1 is that the Kwacha's rate of depreciation this year has exceeded that of comparison currencies, not by a little but by quite a substantial margin. In other words, other currencies are also depreciating, but the Kwacha's rate of depreciation is exceptionally high. This suggests that there might be domestic factors over and above the external ones that are responsible for this. The country is currently facing substantial power outages that are going to negatively impact the economy.  Further, the large and widening fiscal deficit has raised uncertainty about its sustainability and how it's likely to be funded.

The combination of all this is likely preventing new investors from channeling foreign exchange our way and/or causing those with investments in Zambia (stock market, local bond market, local real estate market, etc...) to divest away from Zambia.

Further, there is the business of servicing our external debt obligations. The country's total stock of external debt has risen by about 4 times since 2011 (in September 2011, external debt stock stood at US$1.6bn. As at September 2014, external debt had risen to US$4.7bn. This doesn't include the recently issued US$1.25bn Eurobond). As you all know, external debt service payments have to be made in foreign currency.

Perhaps, the Kwacha's worse than average performance this year is due to the fact that, unlike other Central Banks, the Bank of Zambia (BoZ) has been unwilling to use precious foreign reserves to "defend" the currency. That is, everyone is facing pressure with their currencies but other Central Banks, unlike BoZ, have been throwing fire power at the problem. I doubt this is the case though, precisely because the Central Bank has in actual fact been using reserves to try and stabilize the exchange rate whenever necessary (see the discussion around Figure 3 below).

Another graph that I decided to look at was the long-term performance of the Kwacha, over, say, a 4 to 5 year period. The results of this exercise are presented in Figure 2 below.


                                          Figure 2

                                          Source: Bank of Zambia website

The figure shows the Kwacha/US dollar exchange rate from the first quarter of 2011 (11q1) to the second quarter of 2015 (15q2). What is clear from the figure is that the Kwacha has mostly been depreciating over this period. For instance, the average exchange rate for the first quarter of 2011 was K4.76 to the dollar. By the second quarter of 2015, this had increased to about K7.41 to the dollar.

Countries often hold foreign exchange reserves in the same way that households hold savings - a cushion for a "rainy day". Since we pay for imports in foreign currency, we want to hold a chest of foreign currency so that we can still continue to import things in the event that inflows of foreign currency dwindled to zero. (This is why a country's foreign exchange reserves are often reported in "months of import cover"). But sometimes, a country's Central Bank might use some of this bounty to intervene in the foreign exchange market. If a country's currency is facing pressure, the Central Bank might then draw down its stock of reserves by releasing US dollars into the forex market in exchange for local currency. This immediately increases the supply of dollars and reduces the supply of the local currency and hopefully the exchange rate stabilizes. This is precisely what the Bank of Zambia is reported to have done last week.

But over a longer time horizon, what does Zambia's stock of foreign reserves look like especially when viewed through the lens of a depreciating Kwacha? In Figure 3, I report the quarterly Kwacha/dollar exchange rate from quarter 1 2011 to quarter 1 2015 (essentially data from Figure 2 without q3 2015) along with the country's stock of foreign exchange reserves over the same time period [1]. The exchange rate is represented by the blue bars and reserves by the orange line.



                                   Figure 3

                                Source: Bank of Zambia website


In general, the country's stock of reserves, measured absolutely, has grown over the period covered. For instance, the average stock of reserves for quarter 1 2011 was US$2.1bn. By quarter 1 2015, this had increased to US$2.8bn.[2] But looking at the beginning and end obscures a lot of action in between. Looking at the figure closely, one notices that reserves did grow from US$2.1bn in q1 2011 to  US$3.3bn in q4 2012, only to be drawn down to US$2.4bn at the end of q2 2013 (I have marked this period with an arrow in the figure). This was likely a result of a US$900mn attempt by BoZ to stabilize the currency. A similar story can also be seen for q2 2014 where reserves, which had built up to US$3.5bn, were drawn down to USD2.7bn by q1 2015 (again marked with an arrow). This was likely another attempt (this time worth US$800mn in reserves) by BoZ to stabilize the currency ("likely" being an important word, here). Notice that at both time periods, the draw down in reserves coincided with an uptick in the height of the blue bars (i.e. coincided with a depreciation of the currency).

So if the draw downs in reserves have been a response to a depreciating currency, how would one judge the performance of this strategy?


My parting thoughts are really to say that economists care less about the level of the exchange rate than whether it is stable. We want a stable exchange rate much more than we want a "high" or "low" one. This is because stability is really important for long term planning.  


Footnotes
[1] The Bank of Zambia has not yet reported the stock of foreign exchange reserves for q2 of 2015.

[2] But I don't know whether in "real" terms our stock of reserves has indeed increased. To figure this out, I'd have to compute how much of import cover our reserves covered in q1 2011 against how much they covered in q1 2015. (This would require a little bit more looking around on the CSO website to figure out the quarterly dollar values for imports. Feeling slightly lazy to do that so perhaps somebody else might help?). My priors, however, suggest that we are likely covering less import cover now than before simply because, even though reserves have grown, imports have also likely grown and perhaps at a rate faster than the rate of growth in reserves. But I stand to be corrected.

Some exchange rate charts and observations for Zambia

Zambia's currency, the Kwacha, has taken a bit of a beating these last couple of days. According to the currency converter Oanda.com, the currency opened last week at around K8.59 to the US dollar and closed the week at K9.54 to the dollar, a depreciation rate of around 11% in a single week! Oanda's interbank rate as at yesterday was K9.95 (almost K10) to the dollar.

Some commenters following the local currency's gyrations have made the point that the Kwacha's performance this year is not exclusively a Zambian thing. Other African currencies are also feeling the pressure of China's possible slowdown and/or monetary policy developments in the US.  So I decided to compare the Kwacha's performance against the performance of a selected number of African currencies - chosen because they are usually invoked by many as comparison currencies for the Kwacha. I have also included the performance of the Chilean Peso given that Chile, just like Zambia, is vulnerable to a slow-down in copper demand. (Readers are encouraged to do this comparison exercise for as many currencies as possible). The results of this exercise for the period 31st January 2015 to September 4th 2015 are in Figure 1 below.


                                           Figure 1

                                          Source: Oanda.com


What is clear from Figure 1 is that the Kwacha has depreciated the most when compared to the Naira, Chilean Peso, Kenyan Shilling, Ghanaian Cedi, Rand and Ugandan Shilling. The Kwacha has depreciated by about 50% over the period January 2015 to early September 2015. Only the Ugandan Shilling comes close with a depreciation rate of around 30% over this period.

What is clear from Figure 1 is that the Kwacha's rate of depreciation this year has exceeded that of comparison currencies, not by a little but by quite a substantial margin. In other words, other currencies are also depreciating, but the Kwacha's rate of depreciation is exceptionally high. This suggests that there might be domestic factors over and above the external ones that are responsible for this. The country is currently facing substantial power outages that are going to negatively impact the economy.  Further, the large and widening fiscal deficit has raised uncertainty about its sustainability and how it's likely to be funded.

The combination of all this is likely preventing new investors from channeling foreign exchange our way and/or causing those with investments in Zambia (stock market, local bond market, local real estate market, etc...) to divest away from Zambia.

Further, there is the business of servicing our external debt obligations. The country's total stock of external debt has risen by about 4 times since 2011 (in September 2011, external debt stock stood at US$1.6bn. As at September 2014, external debt had risen to US$4.7bn. This doesn't include the recently issued US$1.25bn Eurobond). As you all know, external debt service payments have to be made in foreign currency.

Perhaps, the Kwacha's worse than average performance this year is due to the fact that, unlike other Central Banks, the Bank of Zambia (BoZ) has been unwilling to use precious foreign reserves to "defend" the currency. That is, everyone is facing pressure with their currencies but other Central Banks, unlike BoZ, have been throwing fire power at the problem. I doubt this is the case though, precisely because the Central Bank has in actual fact been using reserves to try and stabilize the exchange rate whenever necessary (see the discussion around Figure 3 below).

Another graph that I decided to look at was the long-term performance of the Kwacha, over, say, a 4 to 5 year period. The results of this exercise are presented in Figure 2 below.


                                          Figure 2

                                          Source: Bank of Zambia website

The figure shows the Kwacha/US dollar exchange rate from the first quarter of 2011 (11q1) to the second quarter of 2015 (15q2). What is clear from the figure is that the Kwacha has mostly been depreciating over this period. For instance, the average exchange rate for the first quarter of 2011 was K4.76 to the dollar. By the second quarter of 2015, this had increased to about K7.41 to the dollar.

Countries often hold foreign exchange reserves in the same way that households hold savings - a cushion for a "rainy day". Since we pay for imports in foreign currency, we want to hold a chest of foreign currency so that we can still continue to import things in the event that inflows of foreign currency dwindled to zero. (This is why a country's foreign exchange reserves are often reported in "months of import cover"). But sometimes, a country's Central Bank might use some of this bounty to intervene in the foreign exchange market. If a country's currency is facing pressure, the Central Bank might then draw down its stock of reserves by releasing US dollars into the forex market in exchange for local currency. This immediately increases the supply of dollars and reduces the supply of the local currency and hopefully the exchange rate stabilizes. This is precisely what the Bank of Zambia is reported to have done last week.

But over a longer time horizon, what does Zambia's stock of foreign reserves look like especially when viewed through the lens of a depreciating Kwacha? In Figure 3, I report the quarterly Kwacha/dollar exchange rate from quarter 1 2011 to quarter 1 2015 (essentially data from Figure 2 without q3 2015) along with the country's stock of foreign exchange reserves over the same time period [1]. The exchange rate is represented by the blue bars and reserves by the orange line.



                                   Figure 3

                                Source: Bank of Zambia website


In general, the country's stock of reserves, measured absolutely, has grown over the period covered. For instance, the average stock of reserves for quarter 1 2011 was US$2.1bn. By quarter 1 2015, this had increased to US$2.8bn.[2] But looking at the beginning and end obscures a lot of action in between. Looking at the figure closely, one notices that reserves did grow from US$2.1bn in q1 2011 to  US$3.3bn in q4 2012, only to be drawn down to US$2.4bn at the end of q2 2013 (I have marked this period with an arrow in the figure). This was likely a result of a US$900mn attempt by BoZ to stabilize the currency. A similar story can also be seen for q2 2014 where reserves, which had built up to US$3.5bn, were drawn down to USD2.7bn by q1 2015 (again marked with an arrow). This was likely another attempt (this time worth US$800mn in reserves) by BoZ to stabilize the currency ("likely" being an important word, here). Notice that at both time periods, the draw down in reserves coincided with an uptick in the height of the blue bars (i.e. coincided with a depreciation of the currency).

So if the draw downs in reserves have been a response to a depreciating currency, how would one judge the performance of this strategy?


My parting thoughts are really to say that economists care less about the level of the exchange rate than whether it is stable. We want a stable exchange rate much more than we want a "high" or "low" one. This is because stability is really important for long term planning.    


Footnotes
[1] The Bank of Zambia has not yet reported the stock of foreign exchange reserves for q2 of 2015.

[2] But I don't know whether in "real" terms our stock of reserves has indeed increased. To figure this out, I'd have to compute how much of import cover our reserves covered in q1 2011 against how much they covered in q1 2015. (This would require a little bit more looking around on the CSO website to figure out the quarterly dollar values for imports. Feeling slightly lazy to do that so perhaps somebody else might help?). My priors, however, suggest that we are likely covering less import cover now than before simply because, even though reserves have grown, imports have also likely grown and perhaps at a rate faster than the rate of growth in reserves. But I stand to be corrected.