Tactical Asset Allocation + HBPP an intriguing combo

A place to talk about speculative investing ideas for the optional Variable Portfolio

Moderator: Global Moderator

User avatar
InsuranceGuy
Executive Member
Executive Member
Posts: 220
Joined: Sun Mar 29, 2015 1:44 pm

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by InsuranceGuy » Sat Feb 16, 2019 1:59 pm

Kbg -

I agree that the 3 month lookback is diluted when you ensemble 3, 6, and 12 month lookbacks. This appears to be because the 6 and 12 month lookbacks both vote similarly, so they simply outvote the 3 month 2-1. That said, at least ensembling reduces trading volume, while averaging models as you propose would result in 6+ trades/year.

I think ultimately this calls for risk parity or another diversificaton/risk contribution measure as to not equally weight various lookbacks together in a way that both diversifies sufficiently across lookbacks, but doesn't run up trading costs so significantly.

IG
User avatar
ochotona
Executive Member
Executive Member
Posts: 2757
Joined: Fri Jan 30, 2015 5:54 am

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by ochotona » Sat Feb 16, 2019 4:00 pm

Today I'm having a headache even contemplating buying back in possibly into a bull trap on March 1. I might stay put.

The old saying - "when you've won the game stop playing" comes to mind for my situation.
User avatar
Kbg
Executive Member
Executive Member
Posts: 1346
Joined: Fri May 23, 2014 4:18 pm

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by Kbg » Sat Feb 16, 2019 7:58 pm

Well I think doing anything like this is really about trying to hedge on the unknown. That’s the whole point. The method behind 3-6-12 is nothing more complex than doubling the prior and with a pinch of human tendencies to trade off these periodicities.

Tax efficiency aside...if you are trading this system you have to pay attention once a month anyway so the 5mins extra to make the trade doesn’t seem to be too onerous.

If you look closely at the dates of relative outperformance you will see a distinct pattern that I’ve posted about a couple of times here or elsewhere on the board.
User avatar
pugchief
Executive Member
Executive Member
Posts: 2787
Joined: Tue Jun 26, 2012 2:41 pm
Location: suburbs of Chicago, IL

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by pugchief » Sat Feb 16, 2019 9:23 pm

ochotona wrote:
Sat Feb 16, 2019 4:00 pm
Today I'm having a headache even contemplating buying back in possibly into a bull trap on March 1. I might stay put.

The old saying - "when you've won the game stop playing" comes to mind for my situation.
You guys crack me up. The point of the PP is to be agnostic. So you embrace a trading system instead, but then second guess it at every turn so that you aren't even following the automatic rules. 🤔
User avatar
ochotona
Executive Member
Executive Member
Posts: 2757
Joined: Fri Jan 30, 2015 5:54 am

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by ochotona » Sat Feb 16, 2019 9:49 pm

pugchief wrote:
Sat Feb 16, 2019 9:23 pm
ochotona wrote:
Sat Feb 16, 2019 4:00 pm
Today I'm having a headache even contemplating buying back in possibly into a bull trap on March 1. I might stay put.

The old saying - "when you've won the game stop playing" comes to mind for my situation.
You guys crack me up. The point of the PP is to be agnostic. So you embrace a trading system instead, but then second guess it at every turn so that you aren't even following the automatic rules. 🤔
Ahem sir, this the VP forum. Yes, we're pretty damned funny aren't we. We'll figure it all out by end of month, for sure!
User avatar
InsuranceGuy
Executive Member
Executive Member
Posts: 220
Joined: Sun Mar 29, 2015 1:44 pm

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by InsuranceGuy » Sat Feb 16, 2019 10:10 pm

Kbg wrote:
Sat Feb 16, 2019 7:58 pm
Tax efficiency aside...if you are trading this system you have to pay attention once a month anyway so the 5mins extra to make the trade doesn’t seem to be too onerous.
I guess that is my issue, unfortunately 30% of my portfolio is not tax advantaged.

It seems like if tax efficiency is not an issue, why not move to DAA(Defensive)/PAA(Protective)/VAA(Vigilant) or other more reactive and trade heavy method that further reduces drawdowns? I also saw AllocateSmartly has a "Meta" strategy that combines several TAA strategies here:https://allocatesmartly.com/meta-strate ... trategies/.

IG
User avatar
ochotona
Executive Member
Executive Member
Posts: 2757
Joined: Fri Jan 30, 2015 5:54 am

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by ochotona » Sun Feb 17, 2019 7:45 pm

InsuranceGuy wrote:
Sat Feb 16, 2019 10:10 pm
It seems like if tax efficiency is not an issue, why not move to DAA(Defensive)/PAA(Protective)/VAA(Vigilant) or other more reactive and trade heavy method that further reduces drawdowns? I also saw AllocateSmartly has a "Meta" strategy that combines several TAA strategies here:https://allocatesmartly.com/meta-strate ... trategies/.

IG
There is a VAA-P4, not too many assets to trade, sort of a fast, twitchy GEM. Trades a lot. If you had 24 trades a year the friction could be nearly 1% off of your return. 24 * 0.03%. that's not at all funny. That's like giving it up to the ETF providers. Nice theoretical result, tough to do in reality.

GEM trades
Image

VAA-P4 trades
Image
Last edited by ochotona on Mon Feb 18, 2019 11:19 am, edited 2 times in total.
User avatar
ochotona
Executive Member
Executive Member
Posts: 2757
Joined: Fri Jan 30, 2015 5:54 am

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by ochotona » Mon Feb 18, 2019 11:53 am

More pfutzing around. I noticed on Portfoliovisualizers that the 2 and 10 month moving average cross on the S&P500, like on Vanguard VFINX, works very similarly but slightly better than GEM. Backtests to 1987. Also infrequent trades. Your classic "Golden Cross and Death Cross".

NO ONE avoided 1987, that was a cluster****.

Maybe trade 1/2 the portfolio GEM, 1/2 with 2/10 mo MA cross. That's what AlphaArchitect does, but they use the 10 month MA cross itself, not the 2/10 cross itself. The 2/10 cross is better than 10 month MA cross. It gets back to what I was saying... noise cancellation.

How many days until March? Pugchief, at least we're all sitting here in agreement, almost all of the TAA portfolios are in bonds now. No rules have been ignored yet. We're just wondering how to proceed on March 1.

We build the plane as it flies.
User avatar
Kbg
Executive Member
Executive Member
Posts: 1346
Joined: Fri May 23, 2014 4:18 pm

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by Kbg » Mon Feb 18, 2019 1:56 pm

ochotona wrote:
Mon Feb 18, 2019 11:53 am
We're just wondering how to proceed on March 1.
A 1000x no. You have rules or you don't.

If you don't have rules you will live by, then just do whatever you feel like doing which is totally ok on an individual freedom level. Just don't expect any backtesting you have done to have any relevance whatsoever to what it is you are doing.

There are an infinite number of things that could be different in the future and likely will be different. What backtesting will do for you is tell you what behavior you can expect under what conditions.
User avatar
Kriegsspiel
Executive Member
Executive Member
Posts: 1751
Joined: Sun Sep 16, 2012 5:28 pm

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by Kriegsspiel » Mon Feb 18, 2019 6:23 pm

ochotona wrote:
Mon Feb 18, 2019 11:53 am
NO ONE avoided 1987, that was a cluster****.
I get what you're saying but I'm pretty sure 1987 is when Taleb's trade paid off and he was a millionaire all of a sudden.
To die, to sleep
To sleep, perchance to dream; ay, there's the rub
For in that sleep of death what dreams may come
User avatar
ochotona
Executive Member
Executive Member
Posts: 2757
Joined: Fri Jan 30, 2015 5:54 am

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by ochotona » Thu Feb 21, 2019 8:55 am

My plan for end of month:

Trade 1/2 of portfolio with GEM 12. Trade 1/2 of portfolio with 2/10 mo moving average crossovers ("Golden Cross", "Death Cross").

Why 2/10 MA Cross? It works really well, about the same as GEM 12 from 1987 to now. AlphaArchitect trades 1/2 12-month lookback, and 1/2 200 day MA. But 200 day MA is a much poorer choice than 2/10 MA Cross, when you test them on PortfolioVisualizer.

You guys have shown that Relative Momentum 3 6 12 is better for selecting between US and International equities. But using Vanguard SP500 VFINX fund back to 1987, the case hasn't been made to me that multiple lookbacks are better for Absolute Momentum (the risk switch). They looked better when I was using SWPPX back to 1999, but adding more data negates that. I wish I had more data. There is some fragility, luck, instability, noise, call it what you will, in which case I think it's wise to invoke Occam's Razor and do the simpler thing, which is what Gary is encouraging... KISS. But I mind less about fragility since 1/2 of the trendfollowing will be done using a completely different methods, the MA crosses. I think this kind of diversification will be good.

And my trendfollowing exposure is being limited by my overall asset allocation view, which is quite conservative. 40% equities max, even if trendfollowed. So 20% to GEM, 20% to MA Crosses. I recently found that Ned Davis' monthly reports also point risk-averse investors to 40% equities at this time (40% - 70% is their possible range). I also took a survey the results of which pointed me to 40%. So I think the construction is correct. 40% trend, 10% gold, 10% cash, 40% bonds.
User avatar
ochotona
Executive Member
Executive Member
Posts: 2757
Joined: Fri Jan 30, 2015 5:54 am

Re: Tactical Asset Allocation + HBPP an intriguing combo

Post by ochotona » Thu Feb 21, 2019 10:38 am

GEM has another challenge. All of these investment geniuses on at the various wirehouses put in their targets for the S&P500 for the end of 2019, and the average of their sixteen prognoses was... ta-da! Simply the all time high. Was published on CNBC on 1/16/19.

But if GEM does not climb steadily to it's all time high by the end of September, it won't even stay invested. Things get easier by end of October, due to 12 month lookback. We all know that wirehouse gurus are usually overly optimistic; they're talking their book. So the S&P has to do better than the wirehouse chimps think it will do, and 90 days before they say it will do it. A challenge!
Post Reply