More of a theory than an idea.All of the various double long products own t-bills to collateralize relatively small futures positions. The funds then pay out the interest taken in from the t-bills (less whatever the fee is).
If we are ever again in a world where t-bills yield 5-6% might it make sense to build a very simple portfolio (half portfolio really, with the other half in cash) of double long products?
This could allow the the investor to capture the market (remember the market goes up close to 3/4 of the time) plus 5%, or whatever t-bills are yielding.First let me address interest rates. Rates are low by historical standards and at some point the US might try to defend the dollar--these points pave the way for higher rates.
Where I think the idea is interesting is that if equity returns continue to be below normal then an extra 5% for people not comfortable with stock, sector or country selection would allow for having a better shot of adding value to their returns.
So what's not to like?
Well the existing crop of double long products target daily movements of the underlying indexes. This means that over longer periods of time the fund may not capture twice the move of anything. In 2007 SSO lagged SPY and if you recall SPY was up a little for that year.
It's not that SSO didn't work, but that the objective is not to do its thing for more than one day. You can look at the various charts I have included on this post and decide for yourself how well, or not, the double long funds captured a double long effect over longer periods of time.
Now that you have looked at the charts and have an opinion about how well they do capture double long over extended periods of time, there is virtually no way the past result can be repeated. What you see on those charts came about from the combination, sequence and magnitude of all the up and down days that occurred during the time charted. Going forward the same combination, sequence and magnitude of all the up and down days will be different.
The combination over the next year or two could yield a result that looks a lot like double long or looks nothing like double long there is no way to know.
So if that is the case then what is the point? The current roster of products are daily only but that does not make the concept any more or less valid but the right product to capture this does not currently exist. There exists a series of leveraged commodity funds whose objective is double long on a monthly basis (I think these are either from PowerShares of DB or maybe they are one in the same on this, anyone willing to look it up and leave a comment would be much appreciated).
If double long on a monthly basis exists for commodities, can equities for the month or the quarter be that far behind? I am not aware of of any of these in the works but it seems logical and were funds like this ever to list and if t-bill rates were ever to go up toward 6% (remember they were about 5% not too long ago) then executing the theory becomes a little more feasible.
13 comments:
The contango on the futures position more than offsets the interest on the T-bills. There's no free lunch, the guy who sells the futures contract wants to be paid the carry on his hedge.
contango on financial futures? there is no cost to store the s&p 500 like with oil so i thought that this is not a big issue.
according to barchart.com september closed friday at 1254.90 and december closed at 1255.70 which is 6/100 of a percent.
obviously just a snapshot of one moment but i don't think contango is a big issue for financial products.
PIMCO uses TIPS as collateral in their Commodity Real Return Fund (PCRIX).
Based on what I've read the double-long daily volatility funds get about 1.4x over a long period. The problem is the expense drag tends to over emphasise the downside.
SSO for example returns exactly twice the return of the SPX minus the interest rate on T-bills-minus the expenses - I've analysed the returns and proved this before. So the proposed half SSO half T-bills will simply return the same as SPY but with a higher expense ratio. No free lunch, and SSO etc. track the index (minus the interest) much better than you are suggesting...
i am specifically not proposing SSO, I do say it does not work with the daily products. i am exploring whether it could work with a monthly or quarterly product and to be clear I don't know.
Stock index futures at fair value price in the cost of carry minus the dividend yield.
Currently the dividend yield and the carry cost are close to a wash. But not at 6%.
Or at least not at current prices for the S&P. :)
Will we see a 6% dividend yield? Could be. What will interest rates be then? High? Low?
Also, I think you need to do a little more math elsewhere. The other problem is so-called "beta creep".
Exaggerated for the sake of emphasis and ease of calculation. Say the index is 1000, and the 2x fund $10.
Next day, index drops 20% to 800, 2x fund correctly drops 40% to $6. Next day, index rallies 25% back to 1000. 2x fund gains 50% to $9. Oops. Fund is now down 10%, index is unchanged.
What about LEAPs as the product you are looking for, Roger?
Just a quibble: market plus 5% should be market plus 2.5%. Only half is invested in t-bills. Still, an intriguing idea!
if this works then all you need is to buy an index future yourself and leave the cash (required by margin) in t-bills.
The notion that the market goes up 70-70% of the time is not very helpful. The market went up 70% of the time over the last decade, yet we are below where we started 10 years ago, not too useful.
GM went up even a higher percentage of the time over the last 6 decades yet it is back to where it was in the 1950's... again, not too helpful for any decision making process.
These products simply don't work the way a lot of people in the blogosphere seem to assume they do.... The reason that the sellers say they match the daily performance better is because 5% or 3% or whatever negative interest and fees per year is very little per day. So the daily moves of the ETF are very close to those of the market. In the longer term that 3-5% p.a. negative alpha adds up and the performance diverges. It isn't true that they have this bizarre percentage daily performance that many here seem to think that have. Just get the data and do the calculations...
Forget the half in 2x and half cash style - as Born2Code said buying a year long future and banking the cash is a better choice as the daily updates of the 2x has an overall negative effect as per Anonymous's 'Exagerated' example.
If the cash earns a rate comparable to the sum of the futures premium and trading costs, which is typically around 2% p.a. above base rate at the present time, then the conventional and futures based approaches compare equally.
When however stock prices decline then some of the cash will be called upon to fund margin calls, and the remaining lower amount of cash has to work harder to achieve overall break-even with that of the conventional approach. When stock prices rise then profits can be taken and a new (higher start priced) futures position opened, so more cash is available and that cash can work less hard to achieve the break-even, or as hard and achieve a degree of out-performance.
Changes or levels of interest rate generally have little bearing as when rates rise so stock prices fall and dividend yields rise which overall generally negate each other. (There may however be exceptions and cases where possible interest rate based arbitrage become apparent).
The big issue however is whether the cash achieves such a 2% above base rate return. For conventional cash deposits that's unlikely so some other more riskier investment has to be sought out for such purpose. When that investment achieves a better rate than base rate + 2% then overall out-performance of the conventional index occurs, and generally it's considered a lot easier to achieve a better than base rate + 2% investment return than it is to outperform the index (I personally use a time diversified stop-loss based style for this - with primary focus upon capital preservation (so funds are available if called upon to fill margin calls) whilst averaging around 60/40 overall stock/cash exposure levels and hence 60/40 stock/cash like investment returns).
If we take this one step further and invested the cash in an index fund, then we'd end up with 100% stock exposure and where that 'cash' return yielded greater than base rate +2% (as might reasonably be expected) then overall out-performance becomes evident. Which sounds ludicrous, so what might be happening? Well to answer that consider that, at least in the UK, since 1869 up to 1999 equities averaged 9.8% whilst T-Bills averaged 4.7%. On this basis T-Bills grew by a factor of 4,585 whilst equities grew by a factor of 189,803. That 5.1% p.a. average spread is well in excess of the 2% p.a. previously identified. Yes the 1999's were an extreme high, but even reducing the 1999 high down by half still leaves a 9.2% average - a 4.2% spread. Such that overall it would appear that equities provide around a 2.2% overall risk-premium benefit.
Conclusions.
1. Historically equities largely undervalued such that rewards have been excessive. More recent decades however have seen more alignment to fairer price levels such that forward returns might be anticipated to be considerably less than that seen historically.
2. Equities are massively overpriced and overdue for a very severe correction (Dow 45 times lower at around 260 levels).
3. Historic risk-premiums continue into the future and as such equities continue to be one of the more rewarding investments and under which arbitrage via futures versus conventional remain evident.
Personally I discount 2) and believe that it will be more a case of a blend of 1) and 3).
Post a Comment