How to maximizing portfolio return not significantly increasing volatility

[yiugn]

I've worked on momentum based modification of HBPP, found it quite practical and expecting far better result to conventional HBPP methodology, even not significantly increasing volatility

I don't know how to upload image file in this forum, so I created blog and uploaded there

Check out this;

http://permanentpp.blogspot.com/

[stone]

yiugn, you conclude by suggesting that using the best three or best two assets might be optimal. Do you have back tests for doing that?

I'm also a bit worried that the optimal look back period from this point forward will prove to be different from the historical optimal look back period. You point out that using a three year look back period causes the momentum advantage to be lost. Perhaps the treadmill is getting faster and in this internet age 12 months is the new three years

[yiugn]

In fact, there's nothing optimal in the trading. I 'never' mean that 6-12 month is all time optimal value, it's just a rough but 'reasonable' number, not a optimized, all time best, holy grail number.

How about this? Why do you rebalance HBPP anually? Is 12 month is a holy grail number which is universally out of overoptimization regardless of any market condition?

How about rebalancing 6 month? or 11 month? aside from the matter of taxation, is 12 month is magical number and golden rule?

Absolutely no, I am 'NEVER' saying that holding 2-3 assets with the highest momentum is all time optimal, all time better than conventional HBPP method, and the best approach.

What I'd like to claim is, excluding the worst performing asset may not be 'optimal' satisfying every investors, but 'REASONABLE' to those expecting higher return not significantly sacrificing risk.

Skeptics may claim like this,

'What if the worst performed asset class performs the best in the next year? don't you think so? and it 'sometimes' really does, then it could underperform conventional HBPP?"

"Therefore, excluding any asset class is 'overoptimization''

I'd like to ask this question.

'What if the worst performed asset class performs the worst in the next year? and it 'QUITE FREQUENTLY' really does, if then can it overperform conventional HBPP in a quite a lot of time?'

'Even although actually the number of outperformance may not be to big, puttting together the Reward/risk ratio combined with 'winning ratio' of worst performed asset class, is there any reason to stubbornly

adhere to include such a trouble maker in a long time horizon, evidenced by countless research and this simple backtest?"

Many people argue like this,

"Hey, look at this, in some time periods it underperforms conventional HBPP, your strategy is a garbage, it loses'

Are we just sticking to such a short term overperformance? Isn't we try to seek for stable return at least 10-20 year scale?

The reason why many people thinks conventional HBPP method holy grail is simple, it's because they judge other strategy taking HBPP as a standard, not because the methodology is really all time perfect.

In some period, when other approach  outperforms HBPP method with lower volatility and even with superior return, they keep silent, and only when they find they underperform HBPP, they cries out,

'Hey look at that, it's underperforming, it's overoptimized'

Is there any holy grail, or any golden rule, any single ultimate best strategy in the trading?

Which is better between 15%/10%, 5%/5% (return/volatility)?

Is there any answer?

What I'd like to say through this approach(including the approach I've previously mentioned) is not this strategy is 'all time superior', and 'all time out of optimization', 'all time better than conventional approach'

It's just another 'reasonable','alternative', and different approach which is worth to take into consideration for those seeking for slightly aggressive return, not taking too much risk.

Yes, of course conventional HBPP is the safest, because it always hold 4 assets everytime, the lowest volatility.

If you are in pursuit of the lowest volatility it would be more reasonable to allocate more money in the cash portion, then you can say, 'Hey, look at this, your conventional HBPP portfolio is more volatile than

my approach!'

Not the best, not the only holy grail, not optimized, but this is a kind of another alternative approach.

As you mentioned, optimal time period for market would vary, which we may not really know, but just setting arbitrary but static number between 6-12 month does not mean that 'it's optimized, posing potential  

risk for ruin by overoptimization' It does not matter for you even to take 1-3 months, not 6-12 month, because there should exist 'short term momentum'

This is evidence by 'M. Faber's approach, and if you backtest with SPY and TLT or IEF in combination rebalanced from 1-12 month, there's not huge difference (Try testing with Ezbacktest software)

Than what about taking 13 months? how about 14.2 months? What about 2.8 month? These question is quite useless.

This is because 'momentum' does exist in the market, even though we may not know the magic number. Just roughly catching up the effect could lead to significant difference.

It's not the matter of the backtest, but the matter of fundamental charanteristics of market, 'market has a trend'

If market has no trend, and completely stochastic, there's no way to make money with trading.

Optimized number range throughout research is not in fact 'all time optimized', but then is it reasonable to take other number like 5 or 10 year?

It does not matter significantly although market cycle changes and deviates from our setting. Out asset class are diversified, aren't they?  aren't we holding at least2 assets?

Should we always find out the best optimized number at all time? It's totally useless and impossible, therefore blaming simply because it's may be true in some market condition is a paradox, rather it's

quite natural.

[stone]

yiugn, I'm not quibbling with your assertion that markets have momentum.  I agree that the evidence that they generally do is solid. I think some people who frequently post on here have a blend of "Decision Moose" momentum holdings and PP and so are kindred spirits of yours. I'm only saying that consistently capturing such momentum is MUCH harder than consistently capturing the unpredictable market meanderings that also certainly occur.

If market has no trend, and completely stochastic, there's no way to make money with trading

That isn't entirely true is it? Even if you had a pure coin flip style random walk model for price movements (such as in the Black Scholes model) there would be a volatility capture rebalancing bonus because it takes a 100% gain to make up a 50% loss.

Ironically randomness is a lot easier to plan for than non-randomness. There are an infinite variety of different types of non-randomness and you never know which type is going to happen next. The random part of price movements is the one certainty that can be relied on and consistently harvested IMO.

With regard to having a more aggressive approach, have you seen the threads on here about having leveraged PP? For me personally, I have no interest in anything beyond trying to preserve purchasing value of our savings but you might find the leveraged PP discussions an interesting comparison to your "momentumPP".
http://gyroscopicinvesting.com/forum/ht ... ic.php?t=3
http://gyroscopicinvesting.com/forum/ht ... ic.php?t=5

[MachineGhost]

Is there any holy grail, or any golden rule, any single ultimate best strategy in the trading?

Which is better between 15%/10%, 5%/5% (return/volatility)?

Is there any answer?

I believe there is.  It is a metric that compares reward to risk and whatever is relatively highest, is superior.  Secondary to that is the CAGR meets an acceptable threshold.  If CAGR does not, then use leverage.

MG

[MediumTex]

yiugn,

Is English not your first language?

I'm hoping the answer is yes, because otherwise I have this creeping feeling that you are some kind of super-sophisticated spambot.

If you are just doing your best to express yourself in a foreign language, then that's cool.  I just want to get a sense of where you are coming from.

***

I have encountered in other forums what seem to be programs that generate forum posts that almost make sense (some of you may know more about this than I do).  One time in another forum I ran across someone who had said in prior posts that he was from the U.S. and had grown up here, but there was a strange quality to all of his posts--something that didn't quite make sense in the syntax.  I actually thought they were just written by someone in a hurry, but just for kicks I asked the poster to share a humorous story from his childhood, since humor is very hard for a bot to pull off convincingly.  What he posted was something like the following:

"Well I could tell you of me and Hamburglar and how we ate all the hamburgers with full mouths and laughing."

When I read that I thought to myself: "whoa."

***

[edit: I clicked on the link to the blog above and I see Asian language characters, so I guess that answers my question.  Welcome to the forum yiugn.]

[l82start]

i just ran his information (the little bit that is public) through a spam prevention that site i use for a forum i admin on, and he came up clean...  the @naver.com domain does have over 150 confirmed spam bots mostly from Korea..   we have a ongoing thread in our moderators section on Turing test bots ..they are definitely getting more tricky to spot, i got one the other day that had managed 16 posts each with a hidden  .gif link that had gotten past everybody, it finally made one to many 1 word posts "that fit" but not exactly, and it made me look deep enough to spot the hidden links...


PS welcome yiugn.. 

[MediumTex]

I've worked on momentum based modification of HBPP, found it quite practical and expecting far better result to conventional HBPP methodology, even not significantly increasing volatility

I'm amazed at such a casually bold statement.

It would be nice if the PP really could be improved upon that easily through some quick backtesting and momentum-based tweaks.

I'm anxious to hear more about the philosophy behind these modifications to HB's basic recipe other than they happen to work when put through a backtesting program.

[clacy]

yiugn,

Is English not your first language?

I'm hoping the answer is yes, because otherwise I have this creeping feeling that you are some kind of super-sophisticated spambot.

If you are just doing your best to express yourself in a foreign language, then that's cool.  I just want to get a sense of where you are coming from.

***

I have encountered in other forums what seem to be programs that generate forum posts that almost make sense (some of you may know more about this than I do).  One time in another forum I ran across someone who had said in prior posts that he was from the U.S. and had grown up here, but there was a strange quality to all of his posts--something that didn't quite make sense in the syntax.  I actually thought they were just written by someone in a hurry, but just for kicks I asked the poster to share a humorous story from his childhood, since humor is very hard for a bot to pull off convincingly.  What he posted was something like the following:

"Well I could tell you of me and Hamburglar and how we ate all the hamburgers with full mouths and laughing."

When I read that I thought to myself: "whoa."

***

[edit: I clicked on the link to the blog above and I see Asian language characters, so I guess that answers my question.  Welcome to the forum yiugn.]

MT, you're wrong. That is actually the funniest thing I've read on the internet in several days.

[MediumTex]

yiugn,

Is English not your first language?

I'm hoping the answer is yes, because otherwise I have this creeping feeling that you are some kind of super-sophisticated spambot.

If you are just doing your best to express yourself in a foreign language, then that's cool.  I just want to get a sense of where you are coming from.

***

I have encountered in other forums what seem to be programs that generate forum posts that almost make sense (some of you may know more about this than I do).  One time in another forum I ran across someone who had said in prior posts that he was from the U.S. and had grown up here, but there was a strange quality to all of his posts--something that didn't quite make sense in the syntax.  I actually thought they were just written by someone in a hurry, but just for kicks I asked the poster to share a humorous story from his childhood, since humor is very hard for a bot to pull off convincingly.  What he posted was something like the following:

"Well I could tell you of me and Hamburglar and how we ate all the hamburgers with full mouths and laughing."

When I read that I thought to myself: "whoa."

***

[edit: I clicked on the link to the blog above and I see Asian language characters, so I guess that answers my question.  Welcome to the forum yiugn.]

MT, you're wrong. That is actually the funniest thing I've read on the internet in several days.

The guy posted what was supposed to be a picture of himself, but it was a group of people who had won a contest put on by a Houston radio station (there was a banner in the background).

He said something like:

"I am in the photo smiling, but I can only tell you that.  You would guess who I am but that wouldn't be what you think."

[MachineGhost]

I've worked on momentum based modification of HBPP, found it quite practical and expecting far better result to conventional HBPP methodology, even not significantly increasing volatility

I'm amazed at such a casually bold statement.

It would be nice if the PP really could be improved upon that easily through some quick backtesting and momentum-based tweaks.

I'm anxious to hear more about the philosophy behind these modifications to HB's basic recipe other than they happen to work when put through a backtesting program.

I did not see that in the interesting results he reported.  Everything had higher risk or lower return and the optimal portfolio was using 3/4 which obviously excludes cash as such would not by definition ever have momentum.  And yet the 3/4 was not identical to EW so I'm left wondering if the cash in 3/4 was different than the cash in EW, or it was weighted 33/33/33.

Momentum strategies may make more return, but they almost always come at a cost of higher risk.

MG

[stone]

What is the motivation behind "pseudo human" spam bots? Who benefits and how?

For what its worth, Yiugn, your posts didn't seem robotic to me and I didn't have trouble following your English. I think a lot of people on here have pondered about how momentum fits into the PP. I just viewed chasing more momentum as not being a low hanging fruit.

[l82start]

What is the motivation behind "pseudo human" spam bots? Who benefits and how?

  the object is to make posts that wont be deleted or spotted as bots, in order to post hidden links or to come back later when they are buried deep under new posts/threads and edit to add links to raise search engine (Google) ranking......    higher search ranking = better advertising dollars and more hits to your site..