Welcome To The February 2015 Surprise Investment Report
In this report we'll take a look at the January effect as its playing out in 2015 and introduce a new model portfolio intended for IRA brokerage accounts.
January Effect
January 2015 was a down month for the S&P500. We've all heard about the January effect in its many different forms - small caps only, the first 5 days, the entire month of January and so forth. We decided to check the relationship between January gains and annual gains each year since 1950.
We chose 1950 since the S&P 500 index is recorded at Yahoo finance since 1950. Actually the present form of the S&P500 dates back to 1957, but we infer that there is a reconstruction from other S&P 500 indices for these additional years.
Here's a histogram of 65 January simple rates of returns.
January Effect
January 2015 was a down month for the S&P500. We've all heard about the January effect in its many different forms - small caps only, the first 5 days, the entire month of January and so forth. We decided to check the relationship between January gains and annual gains each year since 1950.
We chose 1950 since the S&P 500 index is recorded at Yahoo finance since 1950. Actually the present form of the S&P500 dates back to 1957, but we infer that there is a reconstruction from other S&P 500 indices for these additional years.
Here's a histogram of 65 January simple rates of returns.
Here is a summary of the January percentage gains over the 65 years.
Minimum | 1st Quartile | Median | Mean | 3rd Quartile | Maximum |
---|---|---|---|---|---|
-11.32% | -2.74% | 2.10% | 1.15% | 3.98% | 12.16% |
The table above shows the the expected gain for January is 1.15%. Historically, half of the 'Januarys' had a gain is less than 2.1% and half had gains greater than 2.1%. The standard deviation of January gains is 5%, which gives us little confidence about that expected gain of 1.15%. In fact we can say that "we are 50% confident that a January gain will lie between -2.2% and 4.5%."
The annual gain rates are shown in the following histogram.
The annual gain rates are shown in the following histogram.
Some annual statistics are included in the table below. The table indicates that the mean annual gain for the S&P500 is 9.78%.
We can say that "we are 50% confident that an annual (January through December) gain will lie between -1.5% and 21.1%.
We can say that "we are 50% confident that an annual (January through December) gain will lie between -1.5% and 21.1%.
Minimum | 1st Quartile | Median | Mean | 3rd Quartile | Maximum |
---|---|---|---|---|---|
-36.11% | 0.82% | 11.54% | 9.78% | 21.90% | 47.29% |
But the big question is whether there is a relationship between January gains and annual gains. Here's the best fit line between the gains. You can see that the line does not fit the data well as is thus not a good predictor.
The R squared measure of fit is .24 which means that 'statistically', at least this version of the "January effect' the is poor. However if this statistically best predictor was used regardless, then the annual gain for 2015 would be 3.1%.
New Surprise Report Research
Hopefully the last two Surprise Reports that described our new Fidelity Investment Portfolio methodology, have you wondering how the Surprise Investment Portfolio might evolve based on our new methods and back-tests! Check the FIP tab for a summary of the new methodology.
Unlike the Fidelity Investment Portfolio, the SIP was developed for individual brokerage accounts with transaction costs and taxes on dividends and capital gains. In pursuit of the ongoing SIP development we'll introduce a model portfolio for IRA, rollover IRA, and Roth IRA accounts that have transaction costs but no taxes on dividends and gains other than at the time of withdrawals. We thus establish this new portfolio as our Retirement Investment Portfolio, RIP.
We'll again deploy a portfolio with three sub-portfolios. The objectives will be to achieve balance, a high return rate to risk ratio, maintain an ease of updating for investors, and minimize but not eliminate transaction costs. But we will not have the objective of avoiding capital gain and income taxes i.e., tax efficiency.
The first sub-portfolio again will contain fixed income, currencies, or floating rate income funds. We still don't care to forecast interest rates or exchange rates. But we will again go with US fixed income ETFs with the goal of outperforming the Barclays Aggregate Bond Index and to provide asset returns that are uncorrelated or anti-correlated with equities. Here's our RIP-A sub-portfolio asset classes and assets.
Hopefully the last two Surprise Reports that described our new Fidelity Investment Portfolio methodology, have you wondering how the Surprise Investment Portfolio might evolve based on our new methods and back-tests! Check the FIP tab for a summary of the new methodology.
Unlike the Fidelity Investment Portfolio, the SIP was developed for individual brokerage accounts with transaction costs and taxes on dividends and capital gains. In pursuit of the ongoing SIP development we'll introduce a model portfolio for IRA, rollover IRA, and Roth IRA accounts that have transaction costs but no taxes on dividends and gains other than at the time of withdrawals. We thus establish this new portfolio as our Retirement Investment Portfolio, RIP.
We'll again deploy a portfolio with three sub-portfolios. The objectives will be to achieve balance, a high return rate to risk ratio, maintain an ease of updating for investors, and minimize but not eliminate transaction costs. But we will not have the objective of avoiding capital gain and income taxes i.e., tax efficiency.
The first sub-portfolio again will contain fixed income, currencies, or floating rate income funds. We still don't care to forecast interest rates or exchange rates. But we will again go with US fixed income ETFs with the goal of outperforming the Barclays Aggregate Bond Index and to provide asset returns that are uncorrelated or anti-correlated with equities. Here's our RIP-A sub-portfolio asset classes and assets.
Assst Class | Class |
---|---|
US Corporate Bonds - Intermediate | VCIT |
US Government Bonds - Intermediate | VGIT |
The current allocation of bonds in the Barclays bond index proxy fund, LAG. After extensive back-testing, we chose a simple, low cost portfolio - half Vanguard intermediate corporate bonds and half Vanguard US government bonds, VCIT and VGIT. LAG currently has a duration of 5.4 years, while VCIT and VGIT have durations of 7.4 and 5.6 years respectively.
Here's the backtest of RIP-A from the inception of both Vanguard funds in November 2009.
The five year backtest shows the the RIP-A sub-portfolio gained 33% while the bond index gained 25%. However the bond index did have somewhat less volatility. The advantage in gain is most likely due to the higher allocation to slightly longer duration corporate bonds.
Start | End | Trading Days | Calendar Years |
---|---|---|---|
2009-11-23 | 2015-01-30 | 1305 | 5.19 |
Symbol | Arith Mean | Std Dev | Corr | Beta | Ratio | Gain | Geom Mean | Name |
---|---|---|---|---|---|---|---|---|
RIP-A | 5.7% | 4.3% | -0.31 | -0.08 | 1.32 | 1.333 | 5.7% | |
US-BONDS | 4.3% | 3.5% | -0.46 | -0.1 | 1.22 | 1.247 | 4.4% |
Our next sub-portfolio, RIP-B, is again the equity value and dividend portfolio as with the SIP-B. We remain pleased with our five equity value strategies used in in SIP. Here's a summary.
Assst Class | Asset |
---|---|
High Dividend | HDV |
Consumer Staples | VDC |
Low Volatility | SPLV |
Shareholder Yield | SYLD |
Wide Moat | MOAT |
The plot below shows the performance of the RIP-B relative to the S&P500 total return from the inception date of the two bond ETFs in the RIP-A. (Thus we can combine the equity and bond sub-portfolios in the following section.) We also included a highly recommended fund from a top investment advisor, who suggested that his top long term, single recommendation, CWGIX (Capital World Growth and Income Fund,) might challenge the SIP-B. Due to the fund's notable under-performance we checked that the total performance stated on the American Funds site did indeed verified the performance recorded in Yahoo Finance.
The five year backtest statistics are as follows
Start | End | Trading Days | Calendar Years |
---|---|---|---|
2009-11-23 | 2015-01-30 | 1305 | 5.19 |
Symbol | Arith Mean | Std Dev | Corr | Beta | Ratio | Gain | Geom Mean | Name |
---|---|---|---|---|---|---|---|---|
CWGIX | 9.6% | 16.3% | 0.94 | 0.96 | 0.59 | 1.531 | 8.6% | American Funds Capital World Gro & Inc |
RIP-B | 15.1% | 12.7% | 0.93 | 0.75 | 1.18 | 2.091 | 15.3% | |
S&P500 | 14.6% | 15.9% | 1.0 | 1.0 | 0.92 | 1.995 | 14.3% | S&P 500 (Total Return) |
The RIP-B outperformed the S&P 500 on both an absolute and risk adjusted basis.
We backtested a 50 - 50 stock and bond portfolio using RIP-A and RIP-B, denoted RIP-A-B, against a 50 - 50 portfolio of S&P 500 and Barclay bond total return funds, denoted BM. We are very satisfied with the results - during a bull market in US equities and bonds.
We backtested a 50 - 50 stock and bond portfolio using RIP-A and RIP-B, denoted RIP-A-B, against a 50 - 50 portfolio of S&P 500 and Barclay bond total return funds, denoted BM. We are very satisfied with the results - during a bull market in US equities and bonds.
The table below shows that the RIP-A-B outperformed the benchmark index funds on an absolute and risk adjusted basis. The 10.6% compounded annual growth rate for a 50--50 equity and bond portfolio is excellent.
Start | End | Trading Days | Calendar Years |
---|---|---|---|
2009-11-23 | 2015-01-30 | 1305 | 5.19 |
Symbol | Arith Mean | Std Dev | Corr | Beta | Ratio | Gain | Geom Mean | Name |
---|---|---|---|---|---|---|---|---|
RIP-A-B | 10.3% | 6.3% | 0.86 | 0.34 | 1.63 | 1.688 | 10.6% | |
BM | 9.3% | 7.5% | 0.98 | 0.46 | 1.25 | 1.599 | 9.5% |
The third sub-portfolio will be a momentum portfolio consisting of 5 top performing securities. The assets classes are as follows
We backtested the RIP-C portfolio to 2002 when the iShares sector fund were introduced thus making our selection list suitably robust for backtest. Here's the plot.
- iShares or Vanguard S&P sector ETFs
- Energy master limited partnerships
- Real estate investment trusts including the healthcare REITs considered in the SIP-C sub-portfolio
- Gold
- Royalty trusts
We backtested the RIP-C portfolio to 2002 when the iShares sector fund were introduced thus making our selection list suitably robust for backtest. Here's the plot.
So over 13 years, the backtested RIP-C increased 6.8 times and produced a 16% compounded average growth rate, as well as more than double the return rate to risk ratio of the S&P 500.
Start | End | Trading Days | Calendar Years |
---|---|---|---|
2002-02-01 | 2015-02-02 | 3272 | 13.0 |
Symbol | Arith Mean | Std Dev | Corr | Beta | Ratio | Gain | Geom Mean | Name |
---|---|---|---|---|---|---|---|---|
RIP-C | 16.5% | 18.3% | 0.72 | 0.66 | 0.9 | 6.827 | 16.0% | |
S&P500 | 8.4% | 20.1% | 1.0 | 1.0 | 0.42 | 2.275 | 6.5% | S&P 500 (Total Return) |
To compare with the RIP-A and RIP-B sub-portfolios, we ran the RIP-C momentum portfolio from November 2009. Here's the graph.
The performance of the RIP-C was truly excellent. Over the 5.2 year backtest the portfolio achieved a
- 2.46x portfolio value increase,
- compounded annual growth rate of 19%,
- return rate to risk ratio of 1.25x and greater than the S&P500 during an equity bull market, and a
- lower volatility, a lower correlation, and a lower beta risk.
Start | End | Trading Days | Calendar Years |
---|---|---|---|
2009-11-23 | 2015-02-02 | 1306 | 5.19 |
Symbol | Arith Mean | Std Dev | Corr | Beta | Ratio | Gain | Geom Mean | Name |
---|---|---|---|---|---|---|---|---|
RIP-C | 18.5% | 14.8% | 0.78 | 0.72 | 1.25 | 2.458 | 19.0% | |
S&P500 | 14.9% | 15.9% | 1.0 | 1.0 | 0.93 | 2.021 | 14.5% | S&P 500 (Total Return) |
Thanks for reading!