Please Critique My Proposed PP Tweak

[stuper1]

I'm still having trouble with the idea of putting 25% into cash, when I'm in my mid 40s with 20 years to retirement.  The opportunity cost seems too high to me.  Melveyr has written a few times along the lines that cash is a risk knob that you can turn up or down.  I've also read many comments along the lines that the PP is pretty conservative and defensive.  I feel the need to turn up the risk knob a bit and be a little more offensive.  Melveyr wrote somewhere about using cash to pick your risk level and then dividing the other three assets 3x33 with the remaining money.  This was a while back, so I don't know if that is his current thinking, and he did warn about potential drawbacks (e.g., tight money recession).  This was the starting point for my tweak, but my intuition tells me to go a bit heavy on stocks because they have the greatest long-term upside.  I know this probably isn't the best time to be getting into stocks, but the way I look at it, I'm actually getting out of stocks because before I was 60/40 stocks/bonds.  So, here's the plan I'm thinking of:

Stocks/Bonds/Gold/Cash        34/25/25/16

I did some backtesting from 1974 on a couple backtesting websites (riskcog.com and longtermreturns.com), and this allocation in the past would have given an extra annualized 0.5% return over the vanilla PP, without increasing the risk greatly.  I understand that past results can't predict the future, but they should give an idea of reactions to various market conditions.  I'm still trying to decide how to do rebalancing.  Maybe annually, or at +/- 10% bands.

Does anyone have any thoughts about this plan?  Am I crazy to monkey with a good thing?  I feel like getting an extra 0.5% annually over 10+ years is worth it, especially since it doesn't involve any more work on my part.  I'm just using a different asset allocation.  I'm not trying to do market timing or active investing.  I know that I will need to have a bit stronger stomach when the hard times are happening.  When I get closer to retirement, I can turn the risk knob down a bit by increasing my cash percentage.

[MediumTex]

Bear in mind that you are proposing to overweight stocks after one of the strongest cyclical bull markets in history for stocks against the backdrop of what is basically a multi-decade secular bear market for stocks (starting around 2000).

Your proposal would have been great 4 years ago, but 4 years ago everyone would have said you were crazy.  If people tell you that what you are proposing is a good idea today, what does that tell you about what to expect going forward?

Be careful.

[Pointedstick]

MediumTex is right. Right now everyone loves stocks again after a decade of terrible performance. People seem to have forgotten the last big crash and the flat performance of the last full decade. This is just my opinion, but to me the recent performance of the stock market does not really seem to reflect an increase in the underlying productivity in the real economy; it's just money looking for a home after interest rates fell really far, gold started to look toppy, and previously-safe foreign assets (especially in Europe) appeared too risky.

Also keep in mind that cash isn't just a "risk knob"; one of its roles in the portfolio is to help you take advantage of rising interest rates. If (and I don't know if you do) you believe that interest rates will rise in the short-to-medium term, reducing your exposure to the asset class that's best-positioned to take advantage of that seems short-sighted.

If you're going to reduce cash, I think you should load up on more of the other assets in equal quantities and really ask yourself if you could psychologically handle an increase in interest rates, which would devastate bonds and gold and probably be bad for stocks as well.

[melveyr]

I still think of cash as "negative leverage." This isn't unique to me either, that is the mainstream view of MPT and the "portfolio separation theorem" (a term to google if you want to read more about it). Therefore, the amount of optimal cash is largely subjective and it depends upon your willingness and ability to take risk.

However, it is very important to keep in mind that the Fed uses the interest rate on cash as a policy tool. There are times where the Fed explicitly pursues policies that hurt every single investment asset except for cash. To entirely forgo it you must understand that you are fully at their mercy... The tight money recessions usually don't last long after the Fed realizes their mistake, but historically they have been very painful even for the PP.

Finally, I use cash in my PP. I think forgoing cash could make sense for the right investor, but you have to understand what you are getting yourself into especially with regards to tight money recessions.

[MachineGhost]

times along the lines that cash is a risk knob that you can turn up or down.  I've also read many comments along the lines that the PP is pretty conservative and defensive.

I view people that say the PP is pretty conservative and defensive are victims of poor data and/or insufficient history.  Here are the returns and risk for the PP rebalanced annually since 1968:

01/01/68: 7.67% CAR, -22.43% MaxDD
02/01/68: 7.47% CAR, -26.39% MaxDD
03/01/68: 7.39% CAR, -29.65% MaxDD
04/01/68: 7.37% CAR, -30.50% MaxDD
05/01/68: 7.24% CAR, -30.13% MaxDD
06/01/68: 7.23% CAR, -28.68% MaxDD
07/01/68: 7.25% CAR, -27.66% MaxDD
08/01/68: 7.27% CAR, -27.57% MaxDD
09/01/68: 7.29% CAR, -25.37% MaxDD
10/01/68: 7.25% CAR, -24.79% MaxDD
11/01/68: 7.32% CAR, -25.48% MaxDD
12/01/68: 7.38% CAR, -25.00% MaxDD

Since the PP has unequal risk contributions, gold contributes the most risk and so the MaxDD is from the gold bubble popping in 1980.

[buddtholomew]

times along the lines that cash is a risk knob that you can turn up or down.  I've also read many comments along the lines that the PP is pretty conservative and defensive.

I view people that say the PP is pretty conservative and defensive are victims of poor data and/or insufficient history.  Here are the returns and risk for the PP rebalanced annually since 1968:

01/01/68: 8.58% CAR, -15.37% MaxDD
02/01/68: 8.40% CAR, -20.83% MaxDD
03/01/68: 8.31% CAR, -25.10% MaxDD
04/01/68: 8.26% CAR, -26.08% MaxDD
05/01/68: 8.19% CAR, -25.17% MaxDD
06/01/68: 8.19% CAR, -24.37% MaxDD
07/01/68: 8.29% CAR, -23.93% MaxDD
08/01/68: 8.29% CAR, -23.77% MaxDD
09/01/68: 8.30% CAR, -21.94% MaxDD
10/01/68: 8.24% CAR, -20.80% MaxDD
11/01/68: 8.22% CAR, -21.03% MaxDD
12/01/68: 8.27% CAR, -19.71% MaxDD

Since the PP has unequal risk contributions, gold contributes the most risk and so the MaxDD is from the gold bubble popping in 1980.

MG, there seems to be a problem with your dates or perhaps I am misreading the data.

[stuper1]

What I'm trying to do is be agnostic about the future, mindless about the immediate past few years, but mindful about 80 years of history that show that stocks return 10% on average and bonds return 6%.  It's hard to say what gold returns over a long-term average because we only have good data back to 1972; my guess would be something like 6% also, which is why I'm proposing to weight bonds and gold equally.

MT:  you said "after one of the strongest cyclical bull markets in history," but how do you know that it's over?  I'm not saying that it's not, but I really don't know.

To further check on things, I ran some backtests for the period of 2000 to 2009, which was bad for stocks.  My proposed portfolio averaged 6.2%, while the HBPP averaged 6.5%, so I would have lagged a bit, but not too bad.

Compared to where I was previously (60% stocks and 40% intermediate bonds), I've basically lengthened my bond maturity, and traded 25% stocks for gold, so I don't look at this proposal as overweighting stocks, so much as getting out of stocks.

[MediumTex]

What I'm trying to do is be agnostic about the future, mindless about the immediate past few years, but mindful about 80 years of history that show that stocks return 10% on average and bonds return 6%.

I'm assuming those are U.S. stock market returns, as opposed to Rhodesian, Cuban, Lebanese or Argentinian stock returns. 

You never know which countries are going to be the winners and which are going to be losers in these longer term analyses, and thus you have to be careful about the survivorship bias inherent in any returns data that doesn't also factor in all of the capital that was destroyed along the way in those markets that didn't survive.

MT:  you said "after one of the strongest cyclical bull markets in history," but how do you know that it's over?  I'm not saying that it's not, but I really don't know.

I didn't mean "after" to mean it was over.  I meant "after" more like you would say "after that lunch I don't think I will ever be hungry again."  I'm just contrasting the present with the past.

Compared to where I was previously (60% stocks and 40% intermediate bonds), I've basically lengthened my bond maturity, and traded 25% stocks for gold, so I don't look at this proposal as overweighting stocks, so much as getting out of stocks.

The back of the custom van said it best: "If it feels good, do it."

[MachineGhost]

MG, there seems to be a problem with your dates or perhaps I am misreading the data.

Not sure what you mean, it is Month/Day/Year.  Anyway, I just updated the stats as I switched to using another source for the T-Bonds.  Keep in mind that until 1977, it is 20-year bonds but the differences between 20 and 30 are not that great.  There was a discontinuity between my previous 30-year source and FRED...  FRED was a lot higher, so a drop-in merge before 1977 and back to 1973 as in the original stats was not possible.

Also, I switched to using 1yr T-Bills so everyone can see how critical T-Notes are to softening the MaxDD's and helping the return.  I'm actually starting to think that maybe T-Notes are better than a CD ladder as the ability to generate capital gains seems to be really crucial during scary periods.

[MachineGhost]

BTW, as an example of why using annual data that everyone fancies is such B.S., on the exact day that the PP experienced the -30.13% MaxDD, weekly showed -28.93%, monthly and quarterly showed -18.68% and yearly showed -11.90%!!!  As G.I. Joe would say, "knowing is half the battle."

[Pointedstick]

BTW, as an example of why using annual data that everyone fancies is such B.S., on the exact day that the PP experienced the -30.13% MaxDD, weekly showed -28.93%, monthly and quarterly showed -18.68% and yearly showed -11.90%!!!  As G.I. Joe would say, "knowing is half the battle."

Right, but doesn't that sort of assume that you're watching the portfolio every day? If the MaxDD day happens, but you missed it and then the portfolio recovers the next day, what's the big deal? I guess it depends on how long you're willing to wait for the recovery. The PP seems to have a shorter recovery time than most conventional portfolios where it's measured in years.

[MachineGhost]

Right, but doesn't that sort of assume that you're watching the portfolio every day? If the MaxDD day happens, but you missed it and then the portfolio recovers the next day, what's the big deal? I guess it depends on how long you're willing to wait for the recovery. The PP seems to have a shorter recovery time than most conventional portfolios where it's measured in years.

That actually happened to me during the 2008 crisis.  I did not look to see what the MaxDD was while the crisis unfolded until several months into 2009 as I was overconfident in my PP Jr. portfolio.  I was shocked at how large the MaxDD had been and very disturbed at how I narrowly missed a true surprise horror show.  It was due to that drama that actually started me on the path to the PP and all the Jesus-flogging I do to it.

[AgAuMoney]

I view people that say the PP is pretty conservative and defensive are victims of poor data and/or insufficient history.  Here are the returns and risk for the PP rebalanced annually since 1968:

01/01/68: 7.67% CAR, -22.43% MaxDD
02/01/68: 7.47% CAR, -26.39% MaxDD
03/01/68: 7.39% CAR, -29.65% MaxDD
04/01/68: 7.37% CAR, -30.50% MaxDD
05/01/68: 7.24% CAR, -30.13% MaxDD
06/01/68: 7.23% CAR, -28.68% MaxDD
07/01/68: 7.25% CAR, -27.66% MaxDD
08/01/68: 7.27% CAR, -27.57% MaxDD
09/01/68: 7.29% CAR, -25.37% MaxDD
10/01/68: 7.25% CAR, -24.79% MaxDD
11/01/68: 7.32% CAR, -25.48% MaxDD
12/01/68: 7.38% CAR, -25.00% MaxDD

Since the PP has unequal risk contributions, gold contributes the most risk and so the MaxDD is from the gold bubble popping in 1980.

How can that be "since 1968" and the dates be "mm/dd/yy" and incorporate anything in 1980?

[rickb]

I view people that say the PP is pretty conservative and defensive are victims of poor data and/or insufficient history.  Here are the returns and risk for the PP rebalanced annually since 1968:

01/01/68: 7.67% CAR, -22.43% MaxDD
02/01/68: 7.47% CAR, -26.39% MaxDD
03/01/68: 7.39% CAR, -29.65% MaxDD
04/01/68: 7.37% CAR, -30.50% MaxDD
05/01/68: 7.24% CAR, -30.13% MaxDD
06/01/68: 7.23% CAR, -28.68% MaxDD
07/01/68: 7.25% CAR, -27.66% MaxDD
08/01/68: 7.27% CAR, -27.57% MaxDD
09/01/68: 7.29% CAR, -25.37% MaxDD
10/01/68: 7.25% CAR, -24.79% MaxDD
11/01/68: 7.32% CAR, -25.48% MaxDD
12/01/68: 7.38% CAR, -25.00% MaxDD

Since the PP has unequal risk contributions, gold contributes the most risk and so the MaxDD is from the gold bubble popping in 1980.

How can that be "since 1968" and the dates be "mm/dd/yy" and incorporate anything in 1980?

I'm interpreting the dates to be different start dates, and the CAR and MaxDD to be cumulative to now (i.e. these results are what 12 different people would have seen over the past 45 years, starting a PP on the first day of each month in 1968).

MG - Is MaxDD based on daily or monthly data?

[MachineGhost]

I'm interpreting the dates to be different start dates, and the CAR and MaxDD to be cumulative to now (i.e. these results are what 12 different people would have seen over the past 45 years, starting a PP on the first day of each month in 1968).

MG - Is MaxDD based on daily or monthly data?

You got it!  Daily.