Friday, February 3, 2017

Part III Dynamically balanced equity portfolio with a politically corrupt Fed

Return Objectives:

150 % of NGDP with 60% risk of SP500.
6% current
5. The extraordinary nature of the current US Treasury rates and the risks it presents
As the backtesting graph shows, the above would be the most appropriate and best return for least risk portfolio approach. However, the backtest graph shows a sudden loss of value in the last quarter. We feel that in extraordinary times the manager should override the system if it greatly reduces risk and does not greatly increase return. These are extraordinary times and instead of allowing the derived 40% weight in long US Treasurys, the Origin Process has no long US Treasurys. The following explains why.
The Origin process, given the importance of the US Treasury weight, has a complex analysis to the US Treasury rate to ascertain the utility of the US Treasury as an offset.
Currently long US Treasurys will not respond as in the past, but will have a century event jump and a sudden rise in yield (drop in return), that will not be offset by the SP500.
The first step of the analysis is deriving the market pricing of the risk for each additional year in US Treasury maturity. Longer maturity US Treasurys are more risky than shorter with each year of extension. This is commonly called the risk premium, the additional yield the US must pay lenders to issue longer maturity bonds per year of extension.
The risk premium is in the main for inflation risk and the risk of a sudden change in Federal Reserve policy.
Currently the risk premium is extraordinarily low which ostensibly reflects that the market sees no increase in the already low inflation and that there a high likelihood of a Japanese like secular stagnation. This is at odds with the equity market which is perhaps a complete opposite to this stance. Perhaps it is explained by an unusually political Federal Reserve and remaining large impact of the, in Federal Reserve language " extraordinary monetary policy. The conclusion is that long US Treasurys are not useful at this point. Frequent analysis will be able to identify when that changes, which will occur after “normalization” of Federal Reserve policy.
After deriving the risk premium, the Fed Funds rate forwards - this process uses a 7 year in the future Fed Funds rate - and since long US Treasurys in the norm are approximate to NGDP, this forward Fed Funds is labeled a 7 years in the future NGDP. The changes in this NGDP forward rate are noted as well as the current level:
The term structure of the above forward NGDP is also derived which is the market forward view over time, quarter by quarter out to 7 years, of NGDP. The current term structure of NGDP has risen since the election, or since the odds of a Trump win might win, but are still low by a dramatic measure from pre-crisis levels in 2007.
Keeping in mind the term structure changes in forward NGDP, which in other processes is compared to the changes in the US Treasury yield curve, the 7 year NGDP which in the end will be approximate to the long US Treasury yield is graphed versus the US Treasury 20-year maturity. Obviously, the US Treasury has a very large gain in yield to reach normalization levels of around 4% yield and based on how quickly rates rise, in jumps, that rise in rates will occur, when it does, in large jumps. This will also pressure the Federal Reserve to raise rates far more quickly than the market expects. We are overriding the above model of using US Treasurys as an offsets to a business downturn given the potential possibility is too severe a potentially unrewarded risk.

6. Portfolio construction.

This discussion will now walk through the current portfolio build. Broadening the allowed assets from just riskless and risky equity to US Treasurys with the TLT ETF. And the SPDR SP500 sector ETFS. The assets allowed in the portfolio are the most liquid and credit worthy equity available.
Correlation is the starting point. The correlation for short dated is compared to long term correlation.
The current correlation of the six-month run shows the offset of the US Treasurys to SP500 will not be effective so the weights derived are not applied. Currently the US Treasury weight is zero.
The SP500 sectors in the short 6-month time shows the market rally now ongoing is one sector, finance.
Using the correlation and the volatility of the assets, a frontier of the most optimal portfolios is mapped for 6 months, 1 year, 3 year and 5 years efficient frontier.
Currently the portfolio optimized for 6 months of data is “swamped” finance (XLF) and tech (XLK).
The vertical formation below shows the market is currently speculative where active management will not be effective. The maximum cash weight will be used as the US Treasury TLT is systemically rich and too low a yield.
The one year of data shows the market is speculative and justifies a large cash weight. However the flat near vertical sloping to the right shows a curve is starting to form which indicates active management will again shortly be useful.
There are now more sectors providing return than just finance and tech doing so immediately after the election.
The model will be run daily and the cash position will be greatly reduced when a “normal” curve appears.
The two years of data below is vertical from the large negative return in US Treasurys (TLT), as with the one year of data above, a curve will need to form before cash weights are decreased.
The 3 years of data frontier shows the normal efficient frontier shape bowing out allowing return to be maintained yet risk greatly reduced. Active management was effective.
As the two years and less data forms into this shape, then the weights of assets used will be likely close to this three-year data frontier and cash weight eliminated.


The ten-year data does not generate an opinion as much as provides a long-term context for risk management and the potential return and risk. This is also shown in the backtest below that follows the ten-year graph. The ten-year data is the source for the 60/40 traditional “balanced portfolio” which the above shows is not effective. One of the key drivers to successful portfolio managements is the analysis of US Treasury and a dynamic weight between US Treasurys and equity.
Histograms of the above term periods is provided for the least-risk solution and the risky solution, or tangent line.
The first set on the 6 months of data shows what sectors are driving the vertical frontier map given above - finance and tech. And affirms the view that the market is very speculative.
6 Months data histogram of weights for the 6-month frontier optimized portfolio above:
One year of data histogram of weights for the optimized one year of data optimized portfolios:
Two years of data histogram of weights for the optimized two years portfolio above:
During the three-year time span below, the consumer staples sector (XLP) become a keystone weight though finance and tech remain. This suggests that once the high levels of weights prior of the speculative drivers (XLF and XLK) ends in the shorter charts before, the weights used will be finance (XLF), tech (XLK), and staples(XLP).
Three years of data histogram of weights for the optimized portfolio:


Ten years of data histogram of weights for the optimized portfolio:
The above frontiers and optimal weights are put through a backtest for the last ten years. This is not forward looking nor does it predict an outcome, but is plotted to see if there is a coherent structure to the optimized portfolios over time.
Satisfactory results show where the use if US Treasurys and then reweighting to more optimal equity sector mmix can keep pace with the index - the SP500- yet result in about 65% of the index risk.
The change in weights is derived on the backtest portfolio and is noted. Note that in theory the main driver of reducing risk is the use of US Treasurys (TLT) anticipates major market moves.
The “drawdown” of the theoretical backtested portfolio shows that in theory the model recommended weights that would have resulted in 50% of the market loss - the index SP500.
Then the summary of the above in the backtest plot shows the potential for good risk adjusted returns in comparison to the index, the SP500.
A theoretical “what if” is now derived where the backtest is applied to only the SPDR sector ETF equity, with no US Treasurys. The solution derived is 100% equity weights at all times. This shows the power of the US Treasurys during major index downturns of a recession, and that SPDR sector dynamic reweight does reduce risk.
Return and correlation is graphed. The offset of US Treasurys is missing as summed up in the rolling 2 months correlation graph. US Treasurys are not providing offset to US Treasurys.
The correlation grids are repeated.
This affirms the zero weight of US Treasurys.
The realized volatility for the SP500 is graphed for short dated to long term. The vol shows near historical lows which when combined with the vertical frontier graphs above affirm the max cash weights.
US Treasurys vol is mapped in comparison to SP500 vol. The extraordinary low volatility for both assets classes is shown.
The weights applied:
AssetWeights
Cash88%
US Treasurys0%
Equity12%
Equity SectorWeights
Tech - XLK5%
Finance - XLF2%
Industrials - XLI5%
Risk control is heightened during this speculative market with analysis and possible reweighting twice a day.
When a normal distribution of risk is shown, the reweighting will be reduced to once a daily, then weekly.

No comments:

Post a Comment