Anyone going for the Dragon portfolio?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
steve321
Posts: 954
Joined: Sat Sep 09, 2017 9:16 am
Location: Southampton, UK

Anyone going for the Dragon portfolio?

Post by steve321 »

I am becoming more and more convinced that investors who limit themselves to stocks and bonds are victims to recency bias. In a period of structural growth these asset classes do very well, and baby boomers had great returns, but what happens in a time of crisis, when deflation or inflation rear their ugly heads? I have already added a pretty large allocation to gold to my portfolio, and I am very happy with it.

Chris Cole at Artemis tested different portfolios over longer period including the great depression, and came up with the Dragon portfolio which should well in all environments. Besides stocks and bonds, it has gold, commodity trend and long volatility, all in roughly equal proportions.

Anyone using this kind of portfolio? If you do, what do you use to invest in commodity trend and long volatility?
Success does not bring happiness. In fact, happiness IS success. | 'There are only two tragedies in life: one is not getting what one wants, and the other is getting it.' Oscar Wilde
minimalistmarc
Posts: 1636
Joined: Fri Jul 24, 2015 4:38 pm

Re: Anyone going for the Dragon portfolio?

Post by minimalistmarc »

I’m an optimist so I’m just going to stick with equities
dcabler
Posts: 4544
Joined: Wed Feb 19, 2014 10:30 am
Location: TX

Re: Anyone going for the Dragon portfolio?

Post by dcabler »

Has some similarities to Dalio's All-Seasons portfolio: https://portfoliocharts.com/portfolio/a ... portfolio/

Except it also has some very non-bogleheads-ish aspects
Commodities aren't generally well liked around these parts in general, but Dragon takes it a step further and adds market timing to it.
Similar for Long Volatility as described in this link: https://taylorpearson.me/thedragon/#:~: ... all%20risk.

Cheers.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: Anyone going for the Dragon portfolio?

Post by Forester »

steve321 wrote: Sat Oct 10, 2020 4:32 am I am becoming more and more convinced that investors who limit themselves to stocks and bonds are victims to recency bias. In a period of structural growth these asset classes do very well, and baby boomers had great returns, but what happens in a time of crisis, when deflation or inflation rear their ugly heads? I have already added a pretty large allocation to gold to my portfolio, and I am very happy with it.

Chris Cole at Artemis tested different portfolios over longer period including the great depression, and came up with the Dragon portfolio which should well in all environments. Besides stocks and bonds, it has gold, commodity trend and long volatility, all in roughly equal proportions.

Anyone using this kind of portfolio? If you do, what do you use to invest in commodity trend and long volatility?
Mutiny Fund is an "index" of long vol strategies https://mutinyfund.com/ they have their own podcast and have been interviewed many times.

Were this an option I would definitely put 10% to 20% of my portfolio in these strategies, alongside stocks & gold.

I am sceptical of CTA trend following as it's saturated partly due to commodity indexation which kills the premium, instead I would prefer to own more value stocks, gold & gold equity and get commodity exposure that way.
Amateur Self-Taught Senior Macro Strategist
sassyseuss
Posts: 42
Joined: Fri Oct 02, 2020 7:30 pm

Re: Anyone going for the Dragon portfolio?

Post by sassyseuss »

I have a position in silver. Every hedge against trouble is driving down your profits... unless there really is trouble.

Here's a scenario for you: let's say it had been possible to buy a total index fund on January 1st, 1929, and when you retired 30 years later you sold it on December 31st of 1959. How much did your portfolio appreciate, anyone? Well, the answer is roughly 3%. Inflation was low across this period, but it would still have eaten up about half of the interest gains of that 3%.

If you believe there is even the faintest chance we could be entering another Depression, you might be better served to create some small (5% to 10% range) hedges within your fund. If you think that is a logical impossibility that could never happen, then you should ignore commodities, annuities, and most of the other weird corner-case hedges.

As I said, I have a position in silver.
User avatar
nisiprius
Advisory Board
Posts: 52216
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Anyone going for the Dragon portfolio?

Post by nisiprius »

steve321 wrote: Sat Oct 10, 2020 4:32 am...I have already added a pretty large allocation to gold to my portfolio, and I am very happy with it...
Why are you "very happy with it?" Have you held it for long enough to form a judgement?

With regard to the Dragon Portfolio: just as you ask, exactly how is a retail investor supposed to implement the "active long volatility" component of the portfolio? Do you feel competent to do active management? I don't know what real-world fund or ETF we could buy for exposure to active long volatility. There are some hedge funds that say they do it. Indeed :o Artemis runs one, and perhaps this whole thing is a stalking horse for Artemis.

But what about the research? No data or sources are presented for it so we have to take it on faith. But something is odd here:

ETFs that passively invest in volatility", such as VXX were not available before 2009.

The VIX index itself did not exist before 1993.

S&P 500 options themselves did not exist before 1983.

So just how can Artemis know how active long volatility would have performed in 1929?
Last edited by nisiprius on Sat Oct 10, 2020 10:02 am, edited 4 times in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Anyone going for the Dragon portfolio?

Post by JoMoney »

minimalistmarc wrote: Sat Oct 10, 2020 5:12 am I’m an optimist so I’m just going to stick with equities
I tend to be pretty cynical, but try to avoid that with my investing. History has shown that being broadly optimistic about U.S. businesses pays better* :beer

(* exception for people who literally make a living selling fear and high priced portfolios that cater to fear)
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
MarkRoulo
Posts: 1198
Joined: Mon Jun 22, 2015 10:25 am

Re: Anyone going for the Dragon portfolio?

Post by MarkRoulo »

sassyseuss wrote: Sat Oct 10, 2020 9:36 am I have a position in silver. Every hedge against trouble is driving down your profits... unless there really is trouble.

Here's a scenario for you: let's say it had been possible to buy a total index fund on January 1st, 1929, and when you retired 30 years later you sold it on December 31st of 1959. How much did your portfolio appreciate, anyone? Well, the answer is roughly 3%. Inflation was low across this period, but it would still have eaten up about half of the interest gains of that 3%.
That is an ANNUALIZED 3%. Of the S&P 500 (and whatever Schiller uses before it existed) without dividends.

Note: I ran Jan 1 - Jan 1, 1930 to 1960 to get exactly 360 months.

With dividends reinvested the annualized return over the 30 years was 9% (cumulative: 1,200%)

See here: https://dqydj.com/sp-500-return-calculator/
User avatar
nisiprius
Advisory Board
Posts: 52216
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Anyone going for the Dragon portfolio?

Post by nisiprius »

MarkRoulo wrote: Sat Oct 10, 2020 10:00 am
sassyseuss wrote: Sat Oct 10, 2020 9:36 amlet's say it had been possible to buy a total index fund on January 1st, 1929, and when you retired 30 years later you sold it on December 31st of 1959. How much did your portfolio appreciate, anyone? Well, the answer is roughly 3%. Inflation was low across this period, but it would still have eaten up about half of the interest gains of that 3%.
...With dividends reinvested the annualized return over the 30 years was 9% (cumulative: 1,200%)...
And, sassyseuss, we don't need to imagine an index fund or an index. We can look at a real (actively-managed) stock fund running real money. This is the Massachusetts Investors' Trust, MITTX, now part of the MFS fund family. Inception was in 1924, but here are the historical results of a hypothetical investment of $10,000 on 1/1/1929 to 12/31/1959. It underperformed the index (of course), but nevertheless:

Source

Image

You'd have had $92,106.78. Not quite tenfold, darn it, but not bad.
Average (CAGR) 7.68%/year.

Inflation-adjusted, $92,106.78 year-end-1959 dollars = $53,572.31 1929 dollars. Over fivefold in thirty years.
Average (CAGR) 5.75%/year real (inflation-adjusted).

This, again, is for real dollars in a real mutual fund. A real investor could have actually gotten these results.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Anyone going for the Dragon portfolio?

Post by JoMoney »

MarkRoulo wrote: Sat Oct 10, 2020 10:00 am
sassyseuss wrote: Sat Oct 10, 2020 9:36 am I have a position in silver. Every hedge against trouble is driving down your profits... unless there really is trouble.

Here's a scenario for you: let's say it had been possible to buy a total index fund on January 1st, 1929, and when you retired 30 years later you sold it on December 31st of 1959. How much did your portfolio appreciate, anyone? Well, the answer is roughly 3%. Inflation was low across this period, but it would still have eaten up about half of the interest gains of that 3%.
That is an ANNUALIZED 3%. Of the S&P 500 (and whatever Schiller uses before it existed) without dividends.

Note: I ran Jan 1 - Jan 1, 1930 to 1960 to get exactly 360 months.

With dividends reinvested the annualized return over the 30 years was 9% (cumulative: 1,200%)

See here: https://dqydj.com/sp-500-return-calculator/
Inflation adjusted return on US Large Stocks (S&P 500)
1/1/1929 to 1/1/1959 (30 years) was 6.6% CAGR
1/1/1929 to 12/31/1959 (31 years) was 6.7% CAGR

Not inflation adjusted, return on US Large Stocks (S&P 500)
1/1/1929 to 1/1/1959 (30 years) was 8.5% CAGR
1/1/1929 to 12/31/1959 (31 years) was 8.6% CAGR

I might also point out, that looking at individual point-to-point periods can be misleading too. Most people will accumulate by investing over time, and cashing out and spending everything they day you retire isn't a good idea... you might want your retirement portfolio to last another 20 or 30 years after you turn 60 (or whatever age).
The returns you actually get will be an average of the returns that happen over time both in accumulation and in withdrawal.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Anyone going for the Dragon portfolio?

Post by willthrill81 »

steve321 wrote: Sat Oct 10, 2020 4:32 amIf you do, what do you use to invest in commodity trend and long volatility?
Both commodities and long volatility, TMK, have an expected long-term real return of 0%. That's not the case with stocks, so I'm not sure what's to be gained by the commodities and long vol.
The Sensible Steward
snailderby
Posts: 1421
Joined: Thu Jul 26, 2018 11:30 am

Re: Anyone going for the Dragon portfolio?

Post by snailderby »

I haven't carefully read Chris Cole/Artemis's original article, but according to him, what does adding trending commodities and long volatility offer over something like the Permanent Portfolio or All Weather Portfolio?
z3r0c00l
Posts: 3809
Joined: Fri Jul 06, 2012 11:43 am
Location: NYC

Re: Anyone going for the Dragon portfolio?

Post by z3r0c00l »

A strange time period to propose if advocating silver or gold. Silver returned nothing from 1929 - 1959. Sure it didn't fall too much either. (Well it was almost cut in half in just a year from 1929 - 1930 but it recovered quickly.) But lets look at a more recent time period. How did silver and gold do from 1980 - 2000 compared to stocks and bonds?
70% Global Stocks / 30% Bonds
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Anyone going for the Dragon portfolio?

Post by willthrill81 »

z3r0c00l wrote: Sat Oct 10, 2020 10:38 amHow did silver and gold do from 1980 - 2000 compared to stocks and bonds?
Your point is well taken as it would have been difficult for most investors to hold on to gold, which lost a lot during the period you refer to. But I would argue that portfolio performance is far more important than the individual components of a portfolio. Bonds lost to inflation from 1941-1981, but that doesn't mean that bonds are inherently bad.

During the period you reference, gold did poorly, but stocks and bonds did very well. During the next 20 years, gold far outperformed both stocks and bonds. Portfolios that had a reasonable allocation to all three asset classes had a much smoother ride than did those that only included stocks and bonds.
The Sensible Steward
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: Anyone going for the Dragon portfolio?

Post by Uncorrelated »

The backtest used in the article is invalid due to a look-ahead bias, scaling the portfolio volatility ex-post can result in substantially higher risk-adjusted figures for many reasons. This is the same reason inverse volatility looks economically significant, but actually isn't.

I tried to replicate their performance chart. The 60/40 in their backtest underperforms 60/40 by almost two orders of magnitude.

I skimmed their papers, but I can't find any statistical tests or controls for overfitting. No risk based explanation, no nothing.

Same old overfitting story. Nothing to see here. Move along...
000
Posts: 8211
Joined: Thu Jul 23, 2020 12:04 am

Re: Anyone going for the Dragon portfolio?

Post by 000 »

Here's the allocation for those who don't want to scan through the long article:
24% Domestic Equity
18% Fixed Income
21% Active Long Volatility
18% Commodity Trend Following
19% Physical Gold

I have no interest because it appears to be yet another portfolio finely tuned to the backtest rather than based on fundamentals.
NMBob
Posts: 461
Joined: Thu Apr 23, 2015 8:13 pm

Re: Anyone going for the Dragon portfolio?

Post by NMBob »

i guess without volatility part, the risk parity etf - rpar ?

https://rparetf.com/quarterly-reviews/R ... Review.pdf

RPAR invests across four diverse asset classes that are biased to perform well in very different economic environments:
1. Global equities – strong economy; falling inflation
2. Commodities – rising inflation
 Commodity producers – strong economy
 Physical gold – weak economy
3. TIPS – weak economy; rising inflation
4. Treasuries – weak economy; falling inflation/deflation
P4100354
Posts: 12
Joined: Sun Mar 05, 2017 9:32 pm

Re: Anyone going for the Dragon portfolio?

Post by P4100354 »

Also looking into it as well. There are some long vol ETFs that may be an option, such as the TAIL ETF. For your gold allocation, is it physical or an ETF? If the latter, which ETF did you choose?
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: Anyone going for the Dragon portfolio?

Post by Forester »

P4100354 wrote: Sat Oct 10, 2020 6:56 pm Also looking into it as well. There are some long vol ETFs that may be an option, such as the TAIL ETF. For your gold allocation, is it physical or an ETF? If the latter, which ETF did you choose?
One fifth each; global stocks, LT bonds, gold, trend equity, tail puts can be achieved with ETFs. IMHO the etf $ROMO paired with gold would be better than dumping 20% into CTAs, $ROMO has equity tailwind and gold is less expensive to carry.

This ETF portfolio would have sailed through 2008 https://www.portfoliovisualizer.com/bac ... tion5_1=20
Amateur Self-Taught Senior Macro Strategist
heyyou
Posts: 4461
Joined: Tue Feb 20, 2007 3:58 pm

Re: Anyone going for the Dragon portfolio?

Post by heyyou »

Anyone going for the Dragon portfolio?
There is some point where simplicity suits me better, than the complexity of owning what-if asset classes. I expected to adapt my spending to fit my portfolio size, during retirement. From age 55 to age 70, that has been okay, so far. Others are welcome to do whatever suits them.
JackoC
Posts: 4714
Joined: Sun Aug 12, 2018 11:14 am

Re: Anyone going for the Dragon portfolio?

Post by JackoC »

nisiprius wrote: Sat Oct 10, 2020 9:51 am
With regard to the Dragon Portfolio: just as you ask, exactly how is a retail investor supposed to implement the "active long volatility" component of the portfolio? Do you feel competent to do active management? I don't know what real-world fund or ETF we could buy for exposure to active long volatility. There are some hedge funds that say they do it. Indeed :o Artemis runs one, and perhaps this whole thing is a stalking horse for Artemis.

But what about the research? No data or sources are presented for it so we have to take it on faith. But something is odd here:
ETFs that passively invest in volatility", such as VXX were not available before 2009.
The VIX index itself did not exist before 1993.
S&P 500 options themselves did not exist before 1983.
So just how can Artemis know how active long volatility would have performed in 1929?
The link given above
https://taylorpearson.me/thedragon/
has a further link in it to the paper "The Allegory of the Hawk and Serpent" by Cole, the Artemis guy. You click and enter email address to download a copy of the paper (maybe there's a more direct way to get it w/ doing that, but that's what I did). 'Quantitative Notes' pg 42 describes:

-Active Long Volatility (ALV): the numbers they present for long term past history assume a simplified rules based strategy believed to replicate exposure to the Eurekahedge CBOE Long Volatility Hedge Fund Index. The strategy is to seek exposure to what they call SP 500 2.5% Monthly Rolling Put Index when the SP Index has declined >5% over the past 63 trading days, and seek exposure to the SP500 2.5% Monthly Rolling Call Index when it's risen 5%. 'SP500 2.5% Monthly Rolling Put Index' is not a standard term in CBOE literature but appears to refer to the component of various CBOE options indices where you buy a 2.5% out-of-the-money SPX option on a rolling monthly base.

Hence, 'active' does not mean decision based, just that you don't continuously hold options as it says is the implication of 'tail hedging' v ALV (though CBOE's VXTH tail hedge index, using VIX call options, also has no position sometimes based on a rule, level of the VIX in that case). In the body of the paper the strategy is referred to as 'a profoundly simple (and less effective) replication of what ALV managers do for their clients' but interesting it's viewed as good enough to make the point. And the strategy isn't beyond what small investor could do for themselves. I did not see though where it's stated how this is scaled exactly, as in the 'optimal portfolio' was 21% ALV, but it doesn't spell out if that means the notional size of the options is 21% of the total size of your portfolio (equity and not) or something else.

-Implied volatility prior to 1986: Multivariate regression of implied vol 1986-2019 for options on various 'parts of the skew' (IOW various amounts out-of-the-money for calls and puts) against SPX return and realized volatility over preceding 10, 21, 63 trading day period, claiming high R squared. IOW the idea is that implied vol/skew is a fairly predictable function of recent return and realized volatility, so we have a fairly good idea what implied vols would have been prior to the 1980's.

I find this generally plausible, though not beyond dispute. And it's in any case assuming the cumulative knowledge and market risk preferences which have arisen over 1986-2019 always existed. As a historical matter that's debatable even within the known sample: far OTM puts would not be priced where they were prior to the 1987 crash by today's market with same recent realized vol and SPX return. But, since realistically we're using this to look at the future, and distribution of returns and realized vol of the long term past is only being used as example because it's all we have, I don't see it as categorically invalid to synthesize option prices in the past. Mainly, it's in this case like all others using last century or so of US market data a shaky assumption IMO that that short a period of market history in one country in the particular socio-political, geo-political and economic situation of that country is highly relevant to the future. I don't see the synthesized option prices as greatly adding to that general problem.
User avatar
nisiprius
Advisory Board
Posts: 52216
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Anyone going for the Dragon portfolio?

Post by nisiprius »

JackoC wrote: Sun Oct 11, 2020 12:55 pm
nisiprius wrote: Sat Oct 10, 2020 9:51 amWith regard to the Dragon Portfolio: just as you ask, exactly how is a retail investor supposed to implement the "active long volatility" component of the portfolio? Do you feel competent to do active management? I don't know what real-world fund or ETF we could buy for exposure to active long volatility. There are some hedge funds that say they do it. Indeed :o Artemis runs one, and perhaps this whole thing is a stalking horse for Artemis.

But what about the research? No data or sources are presented for it so we have to take it on faith. But something is odd here:
ETFs that passively invest in volatility", such as VXX were not available before 2009.
The VIX index itself did not exist before 1993.
S&P 500 options themselves did not exist before 1983.
So just how can Artemis know how active long volatility would have performed in 1929?
The link given above
https://taylorpearson.me/thedragon/
has a further link in it to the paper "The Allegory of the Hawk and Serpent" by Cole, the Artemis guy. You click and enter email address to download a copy of the paper (maybe there's a more direct way to get it w/ doing that, but that's what I did). 'Quantitative Notes' pg 42 describes... the numbers they present for long term past history assume a simplified rules based strategy believed to replicate exposure to the Eurekahedge CBOE Long Volatility Hedge Fund Index... Hence, 'active' does not mean decision based, just that you don't continuously hold options...
--Implied volatility prior to 1986: Multivariate regression of implied vol 1986-2019 for options on various 'parts of the skew' (IOW various amounts out-of-the-money for calls and puts) against SPX return and realized volatility over preceding 10, 21, 63 trading day period, claiming high R squared. IOW the idea is that implied vol/skew is a fairly predictable function of recent return and realized volatility, so we have a fairly good idea what implied vols would have been prior to the 1980's. I find this generally plausible...
Thanks.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
dml130
Posts: 166
Joined: Wed Sep 02, 2020 11:40 am

Re: Anyone going for the Dragon portfolio?

Post by dml130 »

I skimmed Cole's paper awhile ago. It's an interesting read, but the portfolio strikes me as overly complicated for the typical investor. For example, you essentially have to time the market to use "commodity-trend", if I'm understanding correctly, which to me defeats the purpose of an all-weather type of portfolio. "Long volatility" is another complicated tool, and I think I saw somewhere that cash might be an adequate substitute (correct me if I'm wrong) for what long-vol tries to achieve.
User avatar
Random Musings
Posts: 6772
Joined: Thu Feb 22, 2007 3:24 pm
Location: Pennsylvania

Re: Anyone going for the Dragon portfolio?

Post by Random Musings »

Obviously, this dragon must have some Pixiu in its genes.

RM
I figure the odds be fifty-fifty I just might have something to say. FZ
balbrec2
Posts: 655
Joined: Mon Nov 13, 2017 2:03 pm

Re: Anyone going for the Dragon portfolio?

Post by balbrec2 »

minimalistmarc wrote: Sat Oct 10, 2020 5:12 am I’m an optimist so I’m just going to stick with equities
Total Stock, Total Bond and TIPS covers it for me
JackoC
Posts: 4714
Joined: Sun Aug 12, 2018 11:14 am

Re: Anyone going for the Dragon portfolio?

Post by JackoC »

dml130 wrote: Sun Oct 11, 2020 6:41 pm I skimmed Cole's paper awhile ago. It's an interesting read, but the portfolio strikes me as overly complicated for the typical investor. For example, you essentially have to time the market to use "commodity-trend", if I'm understanding correctly, which to me defeats the purpose of an all-weather type of portfolio. "Long volatility" is another complicated tool, and I think I saw somewhere that cash might be an adequate substitute (correct me if I'm wrong) for what long-vol tries to achieve.
The Harry Browne Permanent Portfolio, a pretty well known idea, is 25% stock, 25% long term bond, 25% T-bill (so basically cash) and 25% gold. The idea is broadly similar to Dragon, aiming to prosper relatively in a very wide variety of market environments including ones where heavily stock, stock/bond portfolios lose a lot of money (though in a sample of relatively few independent 30 yr periods in one country, the US, you 'always' get it back). If the Dragon weights are 24% stock, 18% bond, 19% gold, 21% 'Active Long Volatility' and 18% Commodity Trend Following, the ALV is one of the things replacing the cash. But I don't see how ALV is any kind of an equivalent of cash in terms of performance. The ALV strategy used for the historical analysis in the paper* would often make a lot of money when the market is in a strong trend (you buy options which pay off for more than premiums), would lose money when market volatility is relatively high but there are no sustained trends (you buy options which generally pay off for less than the premium or nothing), and the cash return** when the market is steady and you buy no options per the rule. That would be pretty different than the performance of cash historically in the US.

Others said they suspect the Dragon result is from overfitting the data. Again I think the possible problem with all such analysis actually goes beyond that. 'Overfitting' assumes a stationary statistical distribution of returns and the proposed result just relies too heavily on the particular flukes of what that distribution happened to kick out in the sample period and it won't perform as well out of sample, assuming the same distribution of returns. In the real world nobody is guaranteeing the distribution of returns is even stationary. Although in fairness that also goes for evaluating 60/40 stock/bond based on history. The latter is simpler than Dragon though, as is the Permanent Portfolio.

*as noted in post above, the paper explains that for the purpose of historical analysis it was assumed that 'ALV' was a simple rule to be long 2.5% out-of-the money rolling monthly puts/calls on SPX when SPX has declined/increased 5%+ in the previous 63 trading days, otherwise no position. Although the paper, by a firm offering ALV strategies AIUI, says actual ALV strategies would be more complicated and claims they'd add more value.
**assuming 21% meant you reserved 21% in cash to buy the options.
sassyseuss
Posts: 42
Joined: Fri Oct 02, 2020 7:30 pm

Re: Anyone going for the Dragon portfolio?

Post by sassyseuss »

I seem to have done some bad math earlier, not sure where I went wrong in the Depression-era calculations. Ahh well.
Register44
Posts: 236
Joined: Sun Feb 09, 2020 11:30 am

Re: Anyone going for the Dragon portfolio?

Post by Register44 »

:arrow:
Last edited by Register44 on Mon Jan 11, 2021 7:32 pm, edited 1 time in total.
GaryA505
Posts: 2909
Joined: Wed Feb 08, 2017 1:59 pm
Location: New Mexico

Re: Anyone going for the Dragon portfolio?

Post by GaryA505 »

Register44 wrote: Sat Nov 21, 2020 2:40 pm I would say this is pretty similar to Browne's 4x25 concept. I like Browne's more as it can truly be replicated. The commodity active trading just does not fit in a 100 year portfolio.

Agreeing to skip the commodities. We can take the HB 4x25 and simply swap cash with TAIL. Even without commodities I think this fits with the hawk and the serpent. 50% stocks and bonds serpent, with 50% hawk with gold and long volatility.

Backtests only go back 3 years, but what is cool is the lack of reliance on US stock market correlation with adding volatility.
Original HB4x25 is 59% correlated, which is pretty close to what it has done over longer periods. But by adding TAIL this makes the entire portfolio only 9% correlated with US stocks. Again just over a 3 year period.

https://www.portfoliovisualizer.com/bac ... tion5_2=25

So rather than trying to implement the Dragon, I think considering adding long volatility to the permanent portfolio is the better approach.
I believe that the Harry Browne Permanent Portfolio, the Golden Butterfly, and the Dragon are all variations on the same theme. Of course only the first 2 of those are easy to implement. I like your idea of swapping the cash part of Browne's 4x25 for TAIL. Some people who implement a 4x25 will use short treasuries for the cash part. I have also wondered about using IVOL (TIPS + rates hedge) as part of it.

I believe that TAIL is easy to simulate in PV, but IVOL not so much.
Get most of it right and don't make any big mistakes. All else being equal, simpler is better. Simple is as simple does.
Post Reply