The enduring myth of the collapsing rand

[alfiano]

Hi all

I frequently come across both posts and comments that lament, express concern over or suggest investment decisions based on the supposed common knowledge that the rand has lost massive value and will continue to do so for the foreseeable future.

This is fortunately closer to myth than reality and is based on an easy-to-make misunderstanding of exchange rates.

Yes, the rand depreciates versus the dollar every year. However, this is expected as we have both higher real interest rates and higher inflation, this does not mean the rand has actually lost value. The dollar can also strengthen versus all major currencies, and the rand will weaken (but this has nothing to do with us/our economy/our politics).

Let me give you a few examples, with fictional figures.

Interest rate parity: In the USA in 2022, you can invest $1000 at an interest rate of 5%. The exchange rate is USD/ZAR=10.00. In South Africa, at the same time, you can invest R10000 ($1000) at an interest rate of 10%.

Inflation is assumed to be nil in both countries.

One year later, in 2023: If you invested in the USA, you have $1050. If you invested in South Africa, you have R11000.

If the exchange rate remained the same, you would have $1050 in the USA or $1100 in South Africa. In which case, all US investors would rather invest in South Africa. This can be achieved risk-free using financial instruments that are beyond the scope of this post. The impact of this is that in 2023, $1050 dollars must, all else equal, be equal to R11000.The exchange rate is therefore USD/ZAR 10.48. The rand has depreciated by 4.5%, but you haven't lost any value if your money was earning 10% in the bank.

Purchasing power parity: In the USA in 2022, an iPhone costs $1000. The exchange rate is USD/ZAR=10.00. Therefore an iPhone costs R10000.

You earn R100,000pa.

There is 100% inflation in South Africa, and 0% inflation in the USA.

It is now 2023. You earn R200,000pa (which is worth the same as R100,000 last year), iPhones cost $1000.

If the exchange remained the same, your salary would now buy 20 iPhones, whereas last year it could only buy 10 iPhones. But, that would mean the rand has doubled in strength, which is obviously not the case - South Africa having 100% inflation is not going to cause the rand to strengthen. Therefore, to maintain parity, R200,000 must be able to buy you 10 iPhones. Therefore the exchange rate is USD/ZAR 20.00. The rand has depreciated by 50%, but it has not lost any value.

Dollar strength: In the US in 2022, $100 buys you 3 Taiwanese microchips, 2kg of British cheddar and 1kg of Australian lithium. USD/ZAR is 10.00

After adjusting for inflation between markets, in the US in 2023, $100 buys you 6 Taiwanese microchips, 4kg of British cheddar and 2kg of Australian lithium.

In South Africa in 2022, R1000=$100 buys you 3 microchips, 2kg of cheddar & 1kg of lithium.

In South Africa in 2023, R1000 still buys you 3 microchips, 2kg of cheddar and 1kg of lithium. But R1000 no longer buys you $100, it buys you $50 (USD/ZAR 20). The dollar has strengthened by 100% ie. doubled in value. But this doesn't affect us so much as we only import 9.3% of our imports from the US. So those imports will cost twice as much, but the rest of our imports cost the same.

So, what has happened to the Rand?

Interest rate parity is, to the extent of my knowledge, a more significant driver of exchange rate movements than purchasing power parity. Using 1 September 2023 (or closest available exchange rates - I selected 5 Sept 2023 as it was higher than 1 Sept 2023 and over 10Y), and the St Louis Fed data series (for interest rates, CPI and Real Broad USD index):

Exchange rate


2 September 2003
7.2910

3 September 2013
10.3175

5 September 2023
19.1981

By interest rate parity:

10 year 20 year Depreciation -46.3% -62%
Of which related to interest rates
-20.1%
-42.5%

Of which related to dollar strengthening
-23.1%
-5%

Implied Rand weakening
-12.5%
-30.5%

By purchasing power parity:

10 year 20 year Depreciation -46.3% -62%
Of which related to inflation
-20.7%
-37.6%

Of which related to dollar strengthening
-23.1%
-5%

Implied Rand weakening
-11.8%
-35.9%

Full workings and sources for the 10 year calculations can be examined here.

Okay, so what can we conclude from the above?

The rand has weakened, but not by anywhere near the point of a collapse. If your money was in an interest-bearing account, the rand has lost 1.2%pa in value over ten years, or 1.3%pa over twenty years. Our purchasing power has reduced by 1.1%pa over ten years and 1.5%pa over twenty. Bear in mind that the rand was particularly strong in the mid-2000s, and we've since experienced the GFC, Zuma years/State capture and electricity shortages.

I hear you say:

"Okay, so the rand hasn't collapsed, but look where we are now! Things are going downhill!"

This is not how exchange rates work, the value of the rand today already takes into account our bleak economic outlook, political instability, corruption and electricity shortages. If these things are worse than currently expected, the rand will weaken. If they turn out to be not as bad as expected, the rand will strengthen.

"Okay, so I should be investing in South Africa?"

That's a more complex topic, addressed well by these two videos:

• International diversification
• Emerging markets

(A side consideration to the above videos include that the JSE has an even higher % of offshore revenue than the S&P500 and also major dual-listed companies).

Anyway, what I do hope you take from this is not to give into emotion-driven narratives of the rand collapsing, make sure you properly consider a retirement annuity (which can be up to 45% offshore, and the equity portion can be 60% offshore), which is not for everyone but does get dismissed by some due to the "rand continually losing value".

Do not be afraid of the USD/ZAR sliding. Our real interest rates are currently 1.8% higher than in the US (and our inflation 2% higher), so we would expect the exchange rate to slide by 1.8% per annum.

Limitations:

The figures derived are sensitive to the start and end date, as exchange rates are volatile between days, months and years. For example, if we looked from December 2001 when the exchange rate hit R13.60 to today, we would see that the rand has strengthened massively. That said, the September 2003/2013/2023 figures were not outliers and broadly representative of the average exchange rates in the year.

This analysis also does not mean that individual events/politics/news don't impact exchange rates. Especially in the short term, the rand can devalue/strengthen significantly on a daily/weekly/monthly basis. But over the long term, these individual shocks average out into a picture that we can better analyse.

The above analysis is simplified and doesn't take into account all known factors that impact exchange rates such as trade surplus/deficits.

Please, by all means have a look at my workings, critique my method or analysis, etc, but please don't dismiss it out of hand - exchange rates are by no means simple, if you disagree, make sure you read through those examples and other material carefully first.

Edit 1: As noted in the comments, I made the rather elementary mistake of using nominal rather than real interest rates, this has been fixed. It impacted some of the percentages but ultimately (fortunately) not the conclusions to be drawn.

 

[ne0phyte2021]

@alfiano Good point , as long as our rand salaries keep up with local inflation , but at the moment for me it is not .

 

[alfiano]

@ne0phyte2021 Very true, for many people salaries lag inflation, as noted elsewhere in the thread real salaries have declined in SA in past 5 years. But if you're a minimum wage employee, you've seen real wage growth in the past few decades. In general, wage growth has remained close to inflation.

All that said, being paid less leaves you worse off both in terms of buying SA goods and imported goods. It isn’t directly relevant to rand strength/weakness, if you know what I mean?

 

[parodyofjay]

@ne0phyte2021 Yes as a teacher I get about 30-40% of the salary I'd get in America with about 15% of the cost of living. Can't complain.

 

[deus_ex_machina]

@ne0phyte2021 I'll let you in on a little secret. It never does.

 

[lovelyday10]

@alfiano Excellent post, thanks for taking the time to share a technical view on the topic.

 

[todosan]

@alfiano For the layman it still means imported goods get more expensive.

And as our industry becomes more and more dilapidated more and more gets imported. Thank god our food supply is excellent and food costs remain low.

 

[saidagha]

@todosan Yes, but salaries also tend to rise with inflation, which is higher in South Africa than USA.

 

[deus_ex_machina]

@saidagha Why don't you talk to a Zimbabwean or Argentinean, they will tell you how good a weak currency is...not.

 

[saidagha]

@deus_ex_machina There is a big difference between hyperinflation and a slowly depreciating currency. All currencies depreciate slowly. The USD just depreciates slower than all the others.

 

[deus_ex_machina]

@saidagha The only difference is time, and the cumulative (compounding) effect of inflation. Neither of which is shown in your high falutin calculations

 

[saidagha]

@deus_ex_machina

The only difference is time

Which is a very important difference

and the cumulative (compounding) effect of inflation

Of course, it will be compounding regardless of how slow or fast the change is. Only way to have it not compound is to have no change, which is impossible.

 

[djj1973]

@todosan Imported good get more expensive in nominal not real terms.

 

[davidpeter1]

@todosan Our food costs are not low, and definitely not remaining low. Officially we have the highest or one of the highest food inflation rates in the world, and unofficial representative baskets of household goods track even higher. Certain catagories having been tracking higher than 50% yoy inflation even.

 

[todosan]

@davidpeter1 Tell me you have not been to most other countries without telling me you have not been to most other countries.

 

[davidpeter1]

@todosan I lived in England and Germany for two years each. I’ve traveled through most of Western Europe, large chunk of Southern Africa and a chunk of Asia.

Even in 2018 it already struck me how much food costs as part of income compared to Germany. All my friends who’ve traveled here from England, Sweden, Germany etc in recent years have commented without fail and without prompt on how expensive food is here now. Plus I shared the stats and facts with you. The EU sometimes beats us on food price inflation, but we literally have the highest or second highest food price inflation in the world by official measures.

Tell me you just want to be an ass without telling me you just want to be an ass.

 

[deus_ex_machina]

@todosan He gets it.

 

[faith4l]

@alfiano Good write up.

To add on the retirement part. People need to consider all options they have available to them (and yes hopefully not based on emotion but rather evidence). At the end of the day this will be limited by their knowledge unfortunately.

E.g. how much local bias does one want to have over market representation? Are you comfortable to manage this yourself and rebalance? At lower income is an RA still worth the rebate given reg28 restrictions (equity limit, premium) - what is your withdrawal looking like in retirement and tax considerations? Are you happy to assume reg28 restrictions will remain constant into the future and any change will be for the better? On the other side will CGT even stay constant in future? If CGT dramatically increases, RA might even be favourable at lower tax rates. Is Estate tax something you are concerned about?

If we have a "free lunch" on forex and only pay tax in currency purchased and disposed, so investments in Euro and disposing in Euro and not paying any tax on the currency conversion to rand, how much more is the depreciation worth as assumptions things stay constant vs investing in an emerging currency?

What about fees and options. Internationally you have funds that cost less than SA offering. To keep it simple, will ignore opportunities here like investing for factor exposure and conviction on that. But how much does one save over 20, 30, 40 years with paying less fees (also assuming the offering stays the same), is this really worth it?

So TLDR - there are a lot of considerations that can and should drive one's portfolio (holistically). At some point unfortunately assumptions are going to be made.

 

[alfiano]

@faith4l Absolutely, you make great points. The RA decision is massively complex and differs so much from person to person.

My note on the RA was definitely not saying one must but rather, "if you were dismissing out of some vague notion that rands are going to be worthless, that is a mistake".

Will Reg28 change for the worse? I think the government has shown it takes RA seriously, the two most recent changes were for the better (1/3 early withdrawal & increase in offshore allocation), so I'd suggest the default frame of mind should be no, it's unlikely to. And the default frame of mind on CGT should be yes, it's likely to be increased (last change was 33% to 40%).

There's also the higher expected real returns (which are more relevant to younger investors saving for retirement) on South African equities - 4.57%pa higher than the US per Prof Damodaran. But of course, the concentration risk of having a significant part of your wealth in one (small) market is significant and expected returns vs actual returns vary significantly.

As you say, assumptions have to be made at some point, and for taxpayers in higher brackets, the answer may be to save some inside and some outside. My point really being not all signs point to "avoid".

 

[weeze]

@alfiano Interesting. I can see that the "invisible hand" is at work balancing Forex. Could you give a similar review of the implications within each country, especially with regards to those of us covering debt, rather than having capital to throw at investments. My thinking is that investment cash travels faster than cash for wages. Does this not lead to a more rapid decline in the internal value of the Rand relative to the internal value of the USD. Maybe this is a question about the rate of wage increase relative to the inflation rate.