Analysis of Discovery Capital 200+ (DC200+)
Financial engineering can divide a portfolio up into its parts, and then put it back together again in a new configuration, but the sum of the parts remains the same. Risk cannot be reduced without also reducing expected return.
Assumptions:
These assumptions are off the top of my head, and good enough for a rough analysis. There are ways to derive these values from current market prices to perform a more accurate analysis.
USA Inflation: 2%
RSA Inflation: 6.5%
Real index growth: 2%
Dividend yield: 2%
Standard deviation of annual returns: 20%
Option 1: 70% DJ Eurostoxx50 ETF/30% MSCI USA index ETF (70/30ETF)
Option 2: DC200+
Option 1 expected nominal return per annum: USA inflation+ Inflation differential between USA/RSA (6.5%-2%=4.5%) + Real index growth + Dividend yield = 2%+4.5%+2%+2% = 10.5% p.a.
Option 2 expected nominal return p.a.: USA inflation + Real index growth = 2%+2% = 4%
Expected value of R1 invested after 5 years: Option 1: R1.65 ; Option 2: R1.22
Can you see what the engineers did here?
From what I can find in the small print there are 3 things that are taken away from an investor in option 2 vs option 1 (don’t be alarmed, they do give most of it back in another form as you will see later).
1. The expected depreciation of the Rand vs hard currencies over the next 5 years – which adds to nominal returns earned by investments priced in hard currencies and then translated back into R. The expected annual depreciation of the R vs. $ can be calculated as the difference between the expected inflation rate in RSA vs USA, in my example (6.5%-2%) = 4.5% per annum of return given up in option 2 vs. option 1
2. The dividend yield earned by the stocks underlying the index. ie. 2% p.a. Dividends have value and can be exchanged for something else that has value like downside protection in derivatives markets.
From the brochure :
Dividends from these indices are utilised to provide the enhanced payouts and guarantees at maturity and are therefore not included in the index returns.
3. The investor in option 2 is locked in for 5 years vs option 1. He has to give up liquidity. This is a valuable thing and can be sold for a price in derivatives markets.
Rand depreciation
They refer to an index, not a R or $ value. Their prey (read clients) would assume if they refer to an index composed of US and euro stocks that it must be priced in a hard currency. In fact they are explicitly referring to a Rand index which is evident in this paragraph in the brochure:
Currency Protection
Although the global portfolio is based on offshore markets, it is unaffected by any Rand appreciation or depreciation. You will therefore not be exposed to any risk of currency fluctuations.
See how they put a positive spin on something that is actually expected to impose a 4.5% annual drag on performance.
5 years ago the S&P 500 stood at 879 points, today it is 1960. The index return was 123% over the 5 years. R1 invested turned into R2.23
5 years ago the R/$ was 7.91 vs. 10.66 today. The return over the 5 years for an investor who had a $ claim on the index value would have been 200% in Rand terms. R1 invested turned into R3.
An investor in the DC200+ has a Rand claim on the index so their R1 would now be worth R2.23
The 70/30ETF investor has a $ claim on the index, so their R1 turned into R3. They also have a claim on dividends, if we assume a yield of 2% for the last 5 years, that means R1 turned into R3.32.
Big picture: the expected outcome for R1 invested in option 1 is R1.65 vs R1.22 for option 2 before the goodies that are added back by our engineer to sweeten the deal for option 2 investors.
The table below shows the probability distribution for the 2 options:
Outcome for R1 investment; Option 1 ; Option 2
>R3........................................6%...............1%
>2; <R3................................25%.............10%
=R2.........................................0%.............56%
>R1; <R2...............................57%...............0%
=R1.........................................0%..............30%
>R1; <R0.5.............................12%...............0%
<R0.5.......................................0%...............3%
Look at the tails: > R3 and <R0.5 so 200% gain or 50% loss, the best and worst outcomes on the table. See how option 2 gives up some of the upside, and how the probability of a really bad outcome also increases. So you have a 6% chance to shoot the lights out with option1, but only 1% with option 2 which is a bad thing for DC200+. Also there is zero chance you can lose more than 50% of your money with the 70/30ETF but a whopping 3% chance with DC200+.
3% might not seem like much, but imagine you are retired and dependent on this thing. You could retire at 60 and die at 95, that is seven 5 year periods each time exposed to this 3% risk, so it becomes a 21% (1 in 5) risk of catastrophic retirement ending loss. Remember the 50% loss is in nominal terms, if my inflation assumption of 6.5% pans out the loss actually means your money loses 64% of its buying power after 5 years, and that is without spending a cent of it.
Another negative for DC200+ is that there is a 30% chance that you make zero return, but with 70/30ETF you at least have a 57% chance to make between zero and 100% return.
On the plus side for DC200+: most of the middle of the distribution is very nice:
- The chances that you end up with more than double your money is: 56%+10%+1%=67%; for 70/30ETF it is only: 0%+25%+6% = 31%
- The chances of losing money is reduced from 12% to 3% for DC200+
So the engineer taketh and the engineer giveth.
But what is the value proposition. Does he take more than he gives.
I calculated that the expected outcome for R1 invested ie. All possible outcomes x their probabilities for the 2 options is: Option 1: R1.76 ; option 2: R1.68. So the hidden cost is about 5% upfront.
However the costs for option 1 has not been deducted yet. Investing in these ETFs incurs a cost of .95% per year, which translates to about the same 5% cost as option 2.
So from a value perspective the 2 options are equivalent.
It is for the investor to decide which distribution they find more attractive. I would suggest DC200+ is not good for a highly risk averse investor like a retiree. Bet that was not the impression the brochure gave you.
I think the ideal investor for this product is a young person with a lot of money. Since there are not too many of them around I wonder how those Discovery brokers manage to get so rich - surely they're not selling this to old people.
Also DC200+ has other costs as explained in my earlier post – those costs are a dealbreaker for me.
Also there is tax – on a R100,000 investment DC200+ would be taxed much more heavily than 70/30ETF. If we are talking R5m+ then the tax position for DC200+ starts getting more advantageous than 70/30ETF.
And remember the liquidity and credit risk associated with DC200+
