Reply to Random Roger's "Week of May 13th" - Rainy Day Funds, Just In Case
I was watching Roger Nusbaum's look ahead into next week and his advice really struck a chord with my current feelings. Maybe I'm just stubbornly cautious, but how some people justify large weights for non-speculative portfolios in volatile sectors, countries, or industries is beyond me. I decided to focus on options for each of the categories he mentioned, so here's the break down of the ETFs/CEFs that I'm going to look at.
The following table summarizes a number of parameters. All numbers are calculated with the adjusted close over the intersection of the symbols' ranges to be fair - that is, if one of the funds has traded over 50 sessions only, the funds are only compared over those 50 sessions.
Variance/Mean is what percentage the standard deviation squared is compared to the average price. The returns are calculated over 20 and 100 sessions, although for the infrastructure funds, some of the funds hadn't traded 100 sessions yet. Last is the correlation to the SPY S&P ETF, which should give a good idea of the correlation to the overall market. Note that this is also sorted by increasing Variance/Mean.
For comparison, SPY has a variance/mean of 8% over the last 100 sessions, gained 3.81% in the last 20 sessions, and gained 5.83% in the last 100 sessions.
Edit: I've been reading through more analyst research and prospecti, and I've come to realize that the difference in real coverage and transparency between ETFs and CEFs is greater than I understood it to be. I've also changed the SPY correlation data to weekly log-return price return correlation (aside: if only there were one correlation calculation that were best for all purposes).
| Symbol | Variance/Mean | 20-Session Return | 100-Session Return | Correlation to SPY |
| Convertible | ||||
| GCV | 2.80% | -0.76% | 2.71% | 31.99% |
| NCV | 10.76% | -0.14% | 0.07% | 22.33% |
| NCZ | 11.15% | 0.14% | 0.36% | 15.89% |
| JQC | 18.88% | 1.53% | 6.86% | 14.20% |
| CVF | 20.84% | 3.03% | 7.92% | 10.35% |
| JPC | 21.16% | 2.41% | 6.78% | 13.94% |
| CHY | 21.86% | 1.30% | 4.70% | 9.78% |
| CHI | 30.77% | 0.70% | 5.74% | 20.62% |
| AVK | 43.50% | 0.40% | 2.96% | 32.29% |
| Infrastructure | ||||
| MGU | 6.80% | 1.63% | 62.98% | |
| MFD | 7.24% | 6.87% | -8.74% | |
| GII | 11.66% | 4.16% | 76.82% | |
| Dividend | ||||
| RDR | 3.88% | 1.64% | 2.03% | 71.54% |
| FDL | 4.54% | 3.07% | 1.07% | 92.52% |
| DTN | 7.19% | 2.89% | 6.55% | 85.05% |
| FVD | 10.25% | 3.27% | 8.85% | 89.34% |
| DTD | 11.47% | 2.63% | 6.09% | 79.62% |
| SDY | 15.45% | 3.56% | 5.26% | 75.83% |
| DOO | 16.54% | 3.43% | 6.05% | 83.56% |
| PEY | 20.19% | 3.26% | 5.93% | 78.46% |
| PID | 21.78% | 3.61% | 7.35% | 33.73% |
| DVY | 58.96% | 1.59% | 12.20% | 84.34% |
| Call/Write | ||||
| BEP | 5.12% | 1.14% | 6.73% | 29.00% |
| MCN | 5.33% | 2.49% | 5.34% | 34.14% |
| FFA | 23.41% | -1.28% | 8.59% | 46.64% |
| BEO | 23.99% | 0.09% | 7.09% | 20.83% |
- Michael J Bommarito II's blog
- Add new comment
- 1008 reads
Data source
great stuff studied in your post, can I ask what you use to get the data? Thanks
Thanks, I really appreciate
Thanks, I really appreciate the input. About the data, I've written various automated ways to query data from Yahoo! and a site called Global Financial Data. I can show you them if you're interested, but they unfortunately aren't programs that will work on most people's computers without a good deal of effort.
For the uninatiated can you
For the uninatiated can you tel me what Variance/Mean means. Also if i read your chart correctly.. One of the funds had a 7 % return over 100days with a 25% correlation to the spy. That seems pretty good waht am i missing/
In finance we typically have
In finance we typically have to assume that changes in price are normally distributed. That means they look something like this.
Now, if you want to pick a low-risk fund, you typically want to look for something that is less volatile than your threshold. This typically comes at the cost of maximum possible return. What you want to do is pick something that has the best return for a given volatility level. If you plot all the volatility as the x-axis and the return as the y-axis, you get something like this:
So seeing the variance-mean ratio for a number of related funds gives you an idea how efficient it is at a given risk level.
Also, I think the fund you're referring to is one of the call-write funds, correct? First off, the correlations I'm giving are not very helpful. I'm trying to get them in a more reasonable form but there are a lot of tradeoffs.
About the funds though, the return on these is the extra amount of money made by issuing a call option. It's a little complicated, but unfortunately unless you know specifically what duration options are being bought, it's hard to bet on why they act the way they do. You might want to read a little more on options before trying to grapple with that one though.
Thank you. So the higher the
Thank you. So the higher the varience mean the better? also can you recommend a set of ETF's of other vechiles that offer a faverable volitlity/ return ratio then rogers.
Well, you can really only
Well, you can really only compare two funds' variance-mean ratios if they have the same return, since comparing the variance-mean ratios of a Treasury ETF and an Emerging Market ETF would obviously ignore the fact that they're designed for completely different objectives. That said though, if you have two funds who have the same return (historically calculated return is not equal to actual return!), you want to invest in the one with less volatility, which means it would be a lower variance-mean.
If you want to do more research on this, be careful you don't get variance-mean ratio mixed up with this variance-return frontier. The former is statistical while the latter is due to assumptions in mathematical portfolio theory.