A series of unfortunate events occurred last Thursday/Friday which precluded my ability to review the TSP funds at the critical end-of-month time frame. Given this, we'll move long today.
===================
Summary
- New Allocation as of close on Monday, May 2nd:
C-Fund: 97%
S-Fund: 2%
F-Fund: 1%
- Make the allocation switch by 12:00 EST to be effective 5/2.
===================
TSP Review
With some frustration, the news from Bernanke last Wednesday caused my latest move-to-cash signal to whipsaw, and with Thursday's action, we had a confirmed LONG call across the board. I missed this signal on Thursday/Friday due to events beyond my control, and the weekend review has caught the change in signal.
Allocation into the funds begins with a 100d lookback on daily price changes. I use the actual TSP fund prices, not ETFs. One aspect I am concerned about is risk -- "sell in May and go away" comes to mind, so an analysis of marginal risk contribution is required.
Marginal risk contribution is the anticipated change in volatility if we start with an index (in this case, the S&P500), and then add 1% (arbitrary) of Fund X to the portfolio. The difference in volatility is useful because we can see what the impact of adding each fund will be to daily swings in value, averaged over 100 days:
Here's the results:
- I-Fund: +0.141%/day
- S-Fund: +0.112%/day
- C-Fund: +0.095%/day
- F-Fund: +0.002%/day
As a reference, the S&P500 has a volatility of +0.101/day over the last 100 days.
These results show us that the I-Fund has been extremely volatile as of late, and addition of shares of the I-Fund and S-Fund, because they have an above-average volatility compared to the S&P500, need to be weighed in terms of their returns.
The Sharpe Ratio is a good way to value this, simply because we now know the volatility, and simply need to calculate the returns over the past 100 days. Doing so produces the following Sharpe Ratios, using a risk-free investment of 0.25% annualized:
- I-Fund: 1.96
- S-Fund: 2.14
- C-Fund: 2.20
- F-Fund: 0.18
What these results tell me is that by-and-large, the C-Fund is the best candidate for the majority of our funds, as it has the greatest reward/risk ratio over the past 100 days. Furthermore, the S-Fund comes in a close second, so some portion of the portfolio should most likely contain the S-Fund. The fact that the F-Fund is well under the rest of the funds simply tells me that
Through the magic of math, it's possible to find the combination that produces the best reward/risk ratio of the portfolio above. Surprisingly, I-Fund lands a big, fat goose-egg in terms of allocation -- 0%. This tells me that the marginal contribution of 1% of I-Fund is worse than the combination of some form of S-Fund and F-Fund, hence, we have the stated allocations above.
Here's the target Efficient Frontier, as I know many of you are interested:
From the picture above you can see where the four funds lie on the "perfect" solution curve -- the blue curve.
For you more aggressive, the I-Fund is sitting PERFECTLY on the curve -- but the higher reward is offset by far more risk -- see how far the I-Fund is deviating from the red line? This means that if you invest in the I-Fund, you are taking on a disproportionate amount of risk, e.g., risk is growing faster than gain. This is bad.
At the end of the day, I've selected the allocations shown above:
- C-Fund: 97%
- S-Fund: 2%
- F-Fund: 1%
I don't normally invest in the F-Fund but I think bond yields will be stable here for the next month or so, resulting in an ability to offset the volatility in the S-Fund but still see some upside. Note that over the past 100 days that the F-Fund has a 0.8% gain, so it is somewhat positively correlated with the markets (which is rare in general).
======================
Remember, you are responsible for your own investment decisions, and I am not. Please do your diligence, and please take ownership for your actions.
Regards,
pgd
No comments:
Post a Comment