My Latest Research on the PP's Big Fat Flaw aka Achilles' Heel

[MachineGhost]

I'm just stopping by temporarily to post this.

[align=center][/align]

The blue bars in the above chart reflect a 25% x 3 Vanilla PP weighting and the red bars show how much risk each asset contributes to the overall Vanilla PP at that 25% weighting.  Maybe this will be clearer:

Asset Vanilla Portfolio Weight Vanilla Volatility Contribution to Vanilla Portfolio Risk Vanilla Risk Contribution
Equity 25.00% 16.74% 0.18% 33.42%
Debt 25.00% 12.08% 0.09% 16.65%
Real 25.00% 20.61% 0.26% 49.93%
Portfolio 75.00% 7.24% 0.52% 100.00%
Diversify Ratio 56.90%

I surmise the 49.93% is why many [potential] PPers have a major problem with the 25% gold allocation... they're sensing it via sore experience/gut feel/intuition/fuzzy logic.  These stats are based on historical returns back to 1968.

The Vanilla PP came in third place in the 300-year Monte Carlo asset allocation backtest.  What beat it was a) equalized risk contribution and b) most diversified portfolio.  They both have better risk adjusted returns and slightly less maximum drawdown (roughly -21% vs -24% for the Vanilla PP).  So it appears to me now that the Vanilla PP is like a sloppy version of both equal risk contribution and most diversified portfolio, that was determined mostly empirically during the late 70's to the late 80's instead of strictly quantitatively (understandable, given the technological limitations at the time).  So it's not going to be at all perfect, but a solid foundation to improve upon.

More importantly, there was a bond bear market in the Monte Carlo backtest to the extent of a nominal 40% maximum drawdown.  Now keep in mind that before 1977 I had to use 20-year T-Bonds for the stats, so I think it would be very prudent to wait for a real bear in real-time on 30-year T-Bonds before deciding to increase the bond allocation based on such a relatively short history, Monte Carlo backtest or not.  I note that Ray Dalio of the "All Weather Portfolio" dramatically reduced his T-Bond allocation after suffering substantial portfolio losses after the verbal end of QEternity in 2012.

An interesting aspect about T-Bonds is that the 40-year bear market from 1940-1980 was certainly not severe in nominal terms, as you can see below:

Year T-Bonds Inflation
1940 2.84% 0.71%
1941 1.88% 9.93%
1942 1.74% 9.03%
1943 2.23% 2.96%
1944 2.50% 2.30%
1945 3.36% 2.25%
1946 2.70% 18.13%
1947 0.31% 8.88%
1948 1.22% 2.73%
1949 3.46% -1.83%
1950 0.41% 5.80%
1951 0.45% 5.96%
1952 1.12% 0.91%
1953 3.11% 0.60%
1954 2.98% -0.37%
1955 -1.95% 0.37%
1956 -3.73% 2.83%
1957 4.46% 3.04%
1958 -3.22% 1.76%
1959 -2.34% 1.52%
1960 10.08% 1.36%
1961 1.75% 0.67%
1962 6.12% 1.23%
1963 0.31% 1.65%
1964 4.32% 1.20%
1965 0.02% 1.92%
1966 1.19% 3.36%
1967 -5.09% 3.28%
1968 2.23% 4.71%
1969 -5.11% 5.90%
1970 13.97% 5.57%
1971 9.49% 3.27%
1972 6.46% 3.41%
1973 -7.82% 8.94%
1974 1.16% 12.10%
1975 4.82% 7.13%
1976 17.86% 5.04%
1977 1.65% 6.68%
1978 -1.56% 8.99%
1979 -1.78% 13.25%
1980 -4.98% 12.35%

From a real return perspective, it was an absolute disaster, hitting around an -80% maximum drawdown in real terms.  We don't need to be concerned about that with the PP concept, so long as you hold appropriate real return assets.

Another interesting aspect about T-Bonds, is that the total return to T-Bills since 1928 starting with $100 is virtually neck and neck within the PP:

Year T-Bills T-Bonds
1928 $25.77 $25.81
1929 $28.99 $28.81
1930 $28.72 $29.48
1931 $28.13 $27.47
1932 $24.99 $25.31
1933 $24.49 $26.00
1934 $31.53 $33.27
1935 $32.42 $34.59
1936 $36.76 $38.51
1937 $40.15 $41.35
1938 $36.93 $38.23
1939 $39.94 $41.53
1940 $40.24 $41.37
1941 $39.27 $40.04
1942 $38.63 $39.17
1943 $40.70 $41.45
1944 $43.52 $44.44
1945 $46.13 $47.50
1946 $51.01 $52.19
1947 $50.60 $50.64
1948 $53.25 $53.42
1949 $54.13 $55.35
1950 $56.67 $56.32
1951 $61.28 $60.80
1952 $65.33 $64.99
1953 $68.59 $69.19
1954 $67.55 $68.39
1955 $76.41 $74.32
1956 $83.65 $78.66
1957 $85.68 $86.68
1958 $85.00 $79.78
1959 $93.57 $89.28
1960 $98.93 $103.68
1961 $100.79 $100.38
1962 $108.19 $111.98
1963 $108.22 $105.69
1964 $115.90 $116.89
1965 $123.30 $118.92
1966 $129.06 $125.29
1967 $128.39 $116.10
1968 $135.78 $132.54
1969 $150.59 $135.42
1970 $149.64 $157.14
1971 $152.46 $158.86
1972 $166.89 $171.24
1973 $198.39 $175.00
1974 $233.83 $220.00
1975 $262.53 $255.84
1976 $273.85 $305.33
1977 $298.41 $291.85
1978 $318.55 $297.64
1979 $366.50 $332.10
1980 $518.82 $443.14
1981 $615.95 $528.41
1982 $558.84 $716.18
1983 $660.72 $605.68
1984 $690.98 $725.13
1985 $703.25 $872.25
1986 $837.30 $1,015.18
1987 $976.23 $855.45
1988 $1,042.66 $1,074.51
1989 $1,111.96 $1,238.70
1990 $1,267.40 $1,232.85
1991 $1,281.86 $1,413.12
1992 $1,393.27 $1,436.70
1993 $1,426.45 $1,676.04
1994 $1,600.01 $1,387.38
1995 $1,611.52 $2,051.46
1996 $1,918.20 $1,770.36
1997 $2,009.29 $2,235.15
1998 $2,184.93 $2,462.10
1999 $2,441.46 $2,036.25
2000 $2,542.07 $2,936.65
2001 $2,659.58 $2,619.47
2002 $2,541.27 $2,673.25
2003 $2,606.60 $2,671.29
2004 $2,933.96 $3,150.52
2005 $3,146.74 $3,301.88
2006 $3,454.68 $3,415.10
2007 $3,910.16 $4,110.85
2008 $4,359.24 $5,898.37
2009 $4,327.85 $3,096.55
2010 $4,572.10 $5,002.51
2011 $5,191.52 $6,901.59
h2012 $5,757.16 $5,855.68
2013 $6,133.87 $5,234.66
2014 $5,985.49 $7,640.97

Next up, the historical volatilities since 1968 for stock, bonds and gold in that order:

Trailing Volatility
16.74%
12.08%
20.61%

And the forward-looking (using the "wisdom of the crowds") volatilities as of last week for the next 1.5 years:

Forward Volatility
17.00%
11.82%
19.33%

Gsundheit!

Moving on, here is the equal risk contribution portfolio:

Asset Equal Risk Portfolio Weight Equal Risk Volatility Contribution to Equal Risk Portfolio Risk Equal Risk Contribution
Equity 30.98% 16.74% 0.27% 33.33%
Debt 43.60% 12.08% 0.27% 33.33%
Real 25.42% 20.61% 0.27% 33.33%
Portfolio 100% 8.98% 0.81% 100.00%
Diversify Ratio 58.24%

...rescaled back to the Vanilla PP's level of portfolio volatility (you could set your own target, but why mess with the almost best?):

Asset  Adjusted Vanilla Risk Contribution
Equity 24.97%
Debt 35.14%
Real 20.49%
Cash 19.40%

If you just keep the Vanilla PP equity and debt at 25%, but lower gold to 20.50%, stocks and gold will virtually match in risk contribution.  So a little underweighting of gold and putting the difference into cash practically fixes the Big Fat Flaw:

Asset Vanilla Portfolio Weight Vanilla Volatility Contribution to Vanilla Portfolio Risk Vanilla Risk Contribution
Equity 25.00% 16.74% 0.18% 40.01%
Debt 25.00% 12.08% 0.09% 19.93%
Real 20.50% 20.61% 0.18% 40.06%
Portfolio 70.50% 6.62% 0.44% 100.00%
Diversify Ratio 57.68%

As further evidence this band-aid is superior to the Vanilla PP, the larger the Diversify Ratio (DR) score, the better the assets in the portfolio are diversified and less correlated.  It would actually have better use for diversifying intra-assets or tactical allocation, but that's for another day.

Here's the DR scores for PP portfolios that include cash:

Vanilla PPx4: 43.51%
ERC PPx4: 47.30%
100% into Stock, Debt, Real or Cash: 25%

Here's the ERC PPx4:

Asset Equal Risk Portfolio Weight Equal Risk Volatility Contribution to Equal Risk Portfolio Risk Equal Risk Contribution
Equity 7.67% 16.74% 0.02% 25.00%
Debt 10.80% 12.08% 0.02% 25.00%
Real 6.30% 20.61% 0.02% 25.00%
Cash 75.24% 1.29% 0.02% 25.00%
Portfolio 100% 2.57% 0.07% 100.00%

If we could leverage the cash somehow, we could up the other weights to match the Vanilla PP's volatility:

  Adjusted Vanilla Portfolio Weight
Equity 21.76%
Debt 30.63%
Real 17.86%
Cash 213.42%
Portfolio 283.66%

Here's the Vanilla PP w/cash:

Vanilla Vanilla Portfolio Weight Vanilla Volatility Contribution to Vanilla Portfolio Risk Vanilla Risk Contribution
Equity 25.00% 16.74% 0.18% 33.42%
Debt 25.00% 12.08% 0.09% 16.65%
Real 25.00% 20.61% 0.26% 49.93%
Cash 25.00% 1.29% 0.01% 1.27%
Portfolio 100.00% 7.29% 0.53% 100.00%

So in another 40 years, maybe the murky issue with the T-Bonds can be fixed.  I believe that is more of a duration risk issue than volatility, anyway.  For now, I rest and will knock off 5% from the gold for my strategic allocation weights.  Between this and equalizing the marketcap risk of the equities, I feel there is currently no juice left to squeeze from this rock after three years.  I'm moving on to the VP now.

[Pointedstick]

Nice data, MG. It makes sense to me, and we all know that gold's volatility and risk contribution are the highest, and T-bonds the least. I've toyed with the idea of scaling up the other assets' volatility instead and am considering exchanging the T-bonds for zeroes in one of my PPs for this reason. That raises the risk contribution for bonds to much more closely match that of gold. What's a good stock fund that's like 10 or 15% more volatile than VTI but tracks the same way?

[MachineGhost]

Nice data, MG. It makes sense to me, and we all know that gold's volatility and risk contribution are the highest, and T-bonds the least. I've toyed with the idea of scaling up the other assets' volatility instead and am considering exchanging the T-bonds for zeroes in one of my PPs for this reason. That raises the risk contribution for bonds to much more closely match that of gold. What's a good stock fund that's like 10 or 15% more volatile than VTI but tracks the same way?

There is SSO or SPXL, 2x and 3x leveraged funds respectively.  You could get it to work with a small enough size, see below for 3x.  I actually think the 3x Vanilla PP is damn deadly because of the unequal risk contributions on steroids which is how I came about to this way of research.  I mean, that is almost 150% notional risk contribution from gold.  Nuts!!!!

Adjusted Vanilla Portfolio Weight
Equity 8.32%
Debt 11.71%
Real 6.83%
Cash 73.13%

Since the 3x funds aren't exactly 3x at all times, there's a bit in margin of safety.  Do cash enhancement (I-Bonds, 5-yr CD laddering, 5-yr CD stacking, etc.) and you're sitting very pretty!

[EdwardjK]

"If you just keep the Vanilla PP equity and debt at 25%, but lower gold to 20.50%, stocks and gold will virtually match in risk contribution.  So a little underweighting of gold and putting the difference into cash practically fixes the Big Fat Flaw:"

So by shifting 4.5% of your Gold allocation to cash fixes all "problems"?  That's only $45,000 out of a $1,000,000 portfolio.

Why bother?

[MachineGhost]

"If you just keep the Vanilla PP equity and debt at 25%, but lower gold to 20.50%, stocks and gold will virtually match in risk contribution.  So a little underweighting of gold and putting the difference into cash practically fixes the Big Fat Flaw:"

So by shifting 4.5% of your Gold allocation to cash fixes all "problems"?  That's only $45,000 out of a $1,000,000 portfolio.

Why bother?

No it doesn't fix ALL problems, it fixes the Big Fat Gold Flaw only.  The T-Bonds I don't feel like messing with due to the duration risk and lack of appropriate historical data.  Monte Carlo is only a simulation, not past reality.  And rates could start rising soon in another 40-year bear market which would make this a really incredibly stupid time to switch to a fixed 35%, in hindsight.

In other words, you could be asking yourself during another "Tight Money" why you didn't listen to MG and simply chop off 5% from gold to increase your certainty and make your life more bearable?  Reducing it by 5% reduces the risk contribution by almost 10%.  That is significant.  Heck, this forum turned into a ghost town in 2012 which was nothing compared to "Tight Money".  How about this lovely year:

Stocks T-Bills T-Bonds Gold
1969 -8.24% 5.52% -5.11% -5.75%

Portfolio: -3.4%, Inflation: 5.90%

Or the more infamous 1981:

Stocks T-Bills T-Bonds Gold
1981 -4.70% 16.58% 0.01% -32.60%

Portfolio: -5.18%, Inflation: 8.91%

...which masks a sickening -25% real maximum drawdown.  In a scary situation like that, every little bit helps.  The world was NOT ending.  It was NOT Armageddeon.  Everything was hunky dory with Reagan's and Thatcher's election.  Who wants to be losing that much real wealth in an atmosphere of unbridled optimism?

(Note my numbers are not to compare to Peak2Trough if you want complete accuracy rather than relativity).

And the final answer is, doing so is keeping in with the core PP philosophy of being completely market agnostic by equalizing the risk exposures (beta).  Equal weight simply doesn't do that.  It was a fine compromise for 1987, but this is 2015.

[sophie]

I suspect Harry Browne knew this, but opted to keep the portfolio simple.  Are you advocating a *maximum* of 20% for gold, or a starting value of 20%?  Then where would you put gold rebalance bands, let's say 10-20% with a starting percentage of 15% (I agree with Edwardjk that a shift of 5% is unlikely to do much).  Then cash would start at 35%, with a range of maybe 25-50%.  Does that sound right?

Alternatively, I figured the "excess" gold is useful for backstopping the Boglehead portfolios I'm running in my employer 403b plans, which otherwise would get crushed in a 1970s-style "stagflation". 

[Pfanni]

With all due respect to your research, but one cannot compute markets.

Risk & Value at risk models work... until they don't.

Example: LTCM.

I recommend the book "A random walk down Wall Street".

Let's assume Harry Browne would have come up with an optimal PP weighting of 21.7/18.9/27.3/32.1.
That would have been a tough sell, marketing-wise.
So we can assume he took the 25/25/25/25 weighting as a marketing approach of selling an easy strategy.
Nevertheless with today's data twenty years later the optimal weighting would look slightly different again.

There cannot and there will never be the optimal strategy and asset class weighting. Markets have too many moving parts to compute.
It's kind of the human existence in a nutshell - we have to navigate life with makeshift solutions.

PP equal weighting to me seems the best makeshift solution in a non-computable, non-perfect world.

[Lang]

Well, you could argue that each asset class should be weighted according to its total market cap.

[ochotona]

I just think there won't be one fixed gold allocation number which is going to work for all time periods; the periodicity of gold is too long. Yes, for the data we have at our disposal now, 20% gold maybe looks better in hindsight; but who know what the future holds?

While I absolutely don't advocate short-term market timing, jumping in and jumping out, let's face it: buy-low sell-high never went away. My plan is to sell most of gold, keeping 5%, if it ever goes to multiple thousands of dollars per ounce and tops out and starts going down again (200 day MA cross or other simple tool to keep myself honest), then buy whatever else is undervalued at the time that still has intrinsic value.

And if gold goes to a stupid low price, like $400 / oz, overweight it beyond 25%.

Over a full generation, maybe my average allocation would have been, in hindsight, 15% - 25%. But I am making my weights price sensitive according to a simple rule. All in at $850 / oz or below. Hold 5% at $2000 / oz or more. Linear interpolation. Maybe it shouldn't be linear, but how would I choose what function to use and how to parameterize it? Linear is simple.

Right now, according to the rule, we're at the $5.65 per cup of Starbucks coffee level. When it's $4, time to buy. If it gets to $2.50, buy a lot. But see... even at $5.65 Starbucks level, I still have a significant gold position, just not a full position. I have some protection against short-term Black Swans, but I am not overpaying for the amount of insurance offered. Insurance is not always worth the premium offered at the time! Dental insurance is a good example of that. I am going to wait for an insurance sale.

But my buys and sells would only be done at regularly rebalancings; I don't want to day-trade gold. But sure, if gold (or any of the other assets) has a huge price change over a brief period of time, I'd be looking at the portfolio, as would we all. Otherwise, I just leave it alone!

But only look at inflation adjusted gold prices. Nominal prices will confuse you.

[KevinW]

Asset Vanilla Portfolio Weight Vanilla Volatility Contribution to Vanilla Portfolio Risk Vanilla Risk Contribution
Equity 25.00% 16.74% 0.18% 33.42%
Debt 25.00% 12.08% 0.09% 16.65%
Real 25.00% 20.61% 0.26% 49.93%
Portfolio 75.00% 7.24% 0.52% 100.00%
Diversify Ratio 56.90%

Yes gold's volatility looks high, but also bonds' volatility looks low.

I wonder if this might shed some insight into the long-standing question of whether zero-coupon bonds ought to be used. Could you run this simulation again on a 25x4 PP that uses zeroes for the bonds?

[RobertMa]

Asset Vanilla Portfolio Weight Vanilla Volatility Contribution to Vanilla Portfolio Risk Vanilla Risk Contribution
Equity 25.00% 16.74% 0.18% 33.42%
Debt 25.00% 12.08% 0.09% 16.65%
Real 25.00% 20.61% 0.26% 49.93%
Portfolio 75.00% 7.24% 0.52% 100.00%
Diversify Ratio 56.90%

Yes gold's volatility looks high, but also bonds' volatility looks low.

I wonder if this might shed some insight into the long-standing question of whether zero-coupon bonds ought to be used. Could you run this simulation again on a 25x4 PP that uses zeroes for the bonds?

^ This really interests me. MachineGhost, thanks for your exhaustive work on this and the rest of the forum. I've been using and following the PP for a while now, but this is my first post.

I am thinking that zero coupons could solve part of this problem while freeing up dollars to throw into equities. A equal-risk portfolio of VTI-GLD-EDV plus 25% Cash sounds like my ideal portfolio. MachineGhost, are you able to figure out what portfolio percentages those three non-cash assets should be in order to equalize risk, or tell me how to figure it out myself?

[Tyler]

I take some issue with the title, equating sub-optimal weighting based on historical risk contribution to an Achilles heel.  Even if the PP could theoretically be improved, that doesn't mean the traditional version is dangerously flawed or cause for concern.

That said, I admire the research, MG.  Very interesting.

[stuper1]

Using www.portfoliovisualizer.com, I tried the vanilla PP (VTI/TLT/GLD/SHV rebalanced annually) versus a PP with EDV substituted for TLT, and then a PP with EDV substituted for TLT and with VXF substituted for VTI.  Results are:

1:    7.66% CAGR, 1.15 Sharpe
2:    8.30% CAGR, 1.11 Sharpe
3:    9.02% CAGR, 1.19 Sharpe

This only goes back to 2009 because some of the ETFs didn't exist before then, so not long enough to mean much, but still very interesting.

[MachineGhost]

I suspect Harry Browne knew this, but opted to keep the portfolio simple.  Are you advocating a *maximum* of 20% for gold, or a starting value of 20%?  Then where would you put gold rebalance bands, let's say 10-20% with a starting percentage of 15% (I agree with Edwardjk that a shift of 5% is unlikely to do much).  Then cash would start at 35%, with a range of maybe 25-50%.  Does that sound right?

No, it is a fixed allocation to gold of 20% instead of 25%.  If we assume the rebalancing thresholds are actually +/- 40% then it would be a +/- 8% rebalancing band for gold and cash would be a +/- 12% rebalancing band.

As to using zero coupons...  the low volatility isn't the problem.  The problem is we have no data for 30-year Treasury bonds to ascertain what the actual risk profile would have been during an actual 40-year bear market.  I believe it would be prudent to leave room for an "upside surprise".  In contrast, stocks and gold have been in multiple bull and bear markets.

I will be implementing these new strategic weights to my equalized market cap (EMC) PP which is my new default PP (I looove Schwab!).  I'm still on the fence about the 35% in bonds.  It's not as if you can't ever change it; you can certainly check what the current historical volatility or forward looking volatility is at each rebalance opportunity just in case they shifted higher.  It would take a pretty extreme event to do that to a long-term average though.

The yearly T-Bond volatility for the 1940-1980 bear market was only 6.99%, but this ignores day to day volatility.  Is 17% volatility enough representation?  I need at least a 95% real world confidence level to be confident.  I'm not, even though 6.99% * 1.96 = 13.7004%.  And 99% is 6.99% * 2.576 = 18.00624%.  Call me chicken.

[Cortopassi]

With all due respect to your research, but one cannot compute markets.

Risk & Value at risk models work... until they don't.

Example: LTCM.

I recommend the book "A random walk down Wall Street".

Let's assume Harry Browne would have come up with an optimal PP weighting of 21.7/18.9/27.3/32.1.
That would have been a tough sell, marketing-wise.
So we can assume he took the 25/25/25/25 weighting as a marketing approach of selling an easy strategy.
Nevertheless with today's data twenty years later the optimal weighting would look slightly different again.

There cannot and there will never be the optimal strategy and asset class weighting. Markets have too many moving parts to compute.
It's kind of the human existence in a nutshell - we have to navigate life with makeshift solutions.

PP equal weighting to me seems the best makeshift solution in a non-computable, non-perfect world.

+1.  This would be my position.  Assuming markets are still around in a couple hundred years, just think of the historical analysis that could be done, with any cherry picked or random time interval showing whatever position the analyst is espousing.  What is a "good" enough historical set of data?  1 year?  10 years?  40 years?  I really don't think it matters because tomorrow is unknown.  Greece blows up.  Cell phones are 100% proven to cause brain tumors.  The Apple watch causes a rash.  Another terrorist attack.  Floods.  Locusts.  Real cold fusion.  You get the idea.

I'm not saying 20% or 25% or some other % is right, or trying to question the risk analysis.  All I know is that I evaluate the PP every quarter.  If anything is a few % off the nominal 25%, I will rebalance.  I am counting on the rebalancing to buy low and sell high over time without me trying to predict what should or shouldn't happen.

[Stunt]

I understand what is being proposed and being an engineer, I do have a soft spot for optimizing to maximize returns with limited resources.  Unfortunately investing and the market has almost infinite variables and cannot be thought of like a process with repetitive characteristics and constant inputs.

Your analysis is based on historical information and who is to say bonds will stay low volatility after extraordinary measures by the central banks. The secular bull market in bonds could come to an end and being overweight bonds will lower your returns. Nobody knows the direction of any of the assets and tinkering around the edges adds complexity for no tangible or guaranteed benefit.

The whole point of PP is to have three volatile assets, not the most/least volatile assets but the purest of the assets classes and to consider the portfolio in its entirety. Portfolio returns have more to do with asset mix (stocks, bonds, gold, real estate, commodities, etc) and not weightings and ability to stick with a system without tinkering and timing.

If you want a different weighting based on your comfort level, go for it. Just don't change your framework down the road after bonds get more volatile or gold becomes a less volatile asset.

[MachineGhost]

If you want a different weighting based on your comfort level, go for it. Just don't change your framework down the road after bonds get more volatile or gold becomes a less volatile asset.

Well said!  Some seem to think the 25%x4 was just some weighting pulled from thin air because it was a nice parity, but no, HB did backtest the PP concept over 100-years of history to arrive at that (keep in mind he had to do this with crap data compared to what we have available now).  Granted, he obviously simplified it later on...  but it's in that simplification where the potential errors lie.

Switching the weights constantly would be tactical allocation and I don't even believe this is the proper way to do that.  Tactical is about forcasting the future.  Strategic is about using what worked in the past and sticking to it through thick and thin.  That is all the PP really is.

In my book, equalizing the risk isn't optimizing, it's living up to the philosophy of not forecasting the future which is implicit with equal weighting that is risk contribution unequal.  No, thanks!

[Gabe]

I will be implementing these new strategic weights to my equalized market cap (EMC) PP which is my new default PP (I looove Schwab!).

Your numbers and thinking make a lot of sense to me, so I'm going to follow your lead and reduce the gold allocation. Thanks a ton for sharing this with us. Could you briefly explain or link to a thread where you discuss how you equalize the market cap risk of the equities?

[ochotona]

I think reduced gold is prudent also. I like 15%, but 20% is more attractive to me than 25%, certainly.

[ozzy]

I think reduced gold is prudent also. I like 15%, but 20% is more attractive to me than 25%, certainly.

I use 15% for gold too.

[Gabe]

Ah, found it: http://gyroscopicinvesting.com/forum/st ... component/

[madbean]

I'm sitting at 19% gold after doing my yearly checkup in February and this post confirms my instincts to stay lazy and not tinker too much. Will probably buy a little gold mid-year.

But if we were in another period of gold doing well would we be having this conversation? I suspect not.

[ochotona]

I'm sitting at 19% gold after doing my yearly checkup in February and this post confirms my instincts to stay lazy and not tinker too much. Will probably buy a little gold mid-year.

But if we were in another period of gold doing well would we be having this conversation? I suspect not.

Not on the ascent, but if it gold went back up to multiple $1000s and then started to kick-over like a model rocket with a spent engine and a long parachute delay, then yes we'd still be talking about the gold allocation.

[I Shrugged]

I agree with those who said that 25x4 is just as likely to be as good going forward as any backtested mix will be.  The future is not going to be like the past.  There are a lot of plates being kept spinning now by central banks across the world.  When it all goes down, I suspect you'll wish you had more gold.   That said, if you even have 15 or 20%, you'll be way ahead of most people.

[ochotona]

I agree with those who said that 25x4 is just as likely to be as good going forward as any backtested mix will be.  The future is not going to be like the past.  There are a lot of plates being kept spinning now by central banks across the world.  When it all goes down, I suspect you'll wish you had more gold.   That said, if you even have 15 or 20%, you'll be way ahead of most people.

I will have more gold! But I won't be at 20%-25% until it goes on sale. Global deflation means gold goes on sale.