This is a fair point. But it is also a monumental challenge to figure out what the correct allocation is at the start and at any other given time. Feel free to post advice on this if you want!

Here I just assumed that the 10x10% was a good start and that no new information was provided. I would assume that 10x10% is not essentially different to what many other people here do, and I would also assume that most people's rationalizations for why they choose some specific different numbers, are hard prove that they a better choice.

You should make your asset allocation based on what your own personal goals are. Do you want to take an income from your investments? Do you simply want to grow your money? How much risk do you want to take? How long are you willing to tie capital up for? How liquid do you want to be? etc etc.

This is different for everybody, and risk tolerance varies a lot as do personal circumstances. Most people tend to overestimate their risk tolerance.

It is hard to pinpoint exactly what is "correct" for anyone except to say it is individually specific and will change over time. For some people becoming rich is important, for some simply staying rich (to them) is all that matters.

In the realm of equities (which are not suitable for everyone) a starter point might be to consider owning a global market cap weighted tracker. If you think you can improve on that you must ask yourself why.

This might be the dumbest thing I've ever read. As far as rebalancing goes it makes zero difference what "measuring stick" you use (aside from the calculation work required) because the ratios will be no different. If you were talking about evaluating your returns you'd have a point, but if the majority of your investments are priced in USD, USD the obvious thing to use for rebalancing and anything else unnecessarily complicates things.

I have talked about this before. From very superficial analysis gold should be flat. Which BTW, beats the negative returns associated with holdings fiat in the long term...

1. Yes, it is a huge mess. Personally, I don't think the return on your time is worth it. I would just eye ball it.

10x10 of what? Asset classes.

My opinion is that you are definitely ****ing up by not owning, gold, btc, real estate, and equities.

I think gold is going to get expensive here in a heart beat. Real estate already is, but if you can lock these artificial rates...its free money (unless prices tank, which I can't see long term).

I can't tell if BTC is expensive or not. But I think you've got to have some.

Equities are expensive. And I love the idea of them being the bread a butter of anyone's portfolio. But I would rather have confidence in my average costs than worry about missing out.

I don't like bonds or fiat. But bonds can serve a purpose from a certain perspective.

If I just inherited $1M and had to figure out what to do with it I would buy a good amount of gold and silver. I would get some exposure to BTC. I would buy some income producing real estate w/ a 30 year fixed. And I would take pilot positions in some equities I like just as a marker.

Maybe $50-100K PMs.
$10K BTC.
~20% down on a couple of quality properties might be about $200-250K.

That leaves you with about 60-70% of your capital still left to allocate. But if you just wanted to invest it you could put it in gold, some TIP bond, or leave it in cash / a trading account. Just depends what kind of trigger finger you have and what your investment horizons are like.

I personally think CAMP and such provide a false sense of security, it is built on faulty assumptions. I also think buy and hold is dead unless you were able to buy in '08 or the next cyclical low, etc.

That is also my strategy, but I have this nagging feeling that one day I will regret it :P

This was just a hypothetical assumption of asset classes. But it's not that different from what I actually do. But I go from 2-30% in like ~10 different assets including real estate, stocks, crypto. No metals and very little fiat.

do as you please, you are going to add variance to your picks if you don't know what you are doing.

if you select arbitrary rules like "if it's up 20%, so the weight in the portfolio is too big" you are doing random things so it's just a lucky pick.

ditto if you rebalance every "X months", just because it's cool to rebalance every X months.

you should estimate what is the potential return within time frame X,Y,Z.

reduce your position when the stock is overbought and increase your position when it's oversold.

momentum is one of the major factor that contribute to raise in price, so if a stock is trending higher and for a long time and it's clearly doing better than the others in your protfolio, you shouldn't sell it just to rebalance, see if the thesis is still ok, if time-adjusted and risk-adjusted is still fine to hold,

e.g. if you bought at $50, and it's at $90 within 1 year, while the rest of the portfolio is in the $40-60 range, and the now $90 stock has almost reach a plateaux and it won't go much higher than $100 in the next year, while the rest of your portfolio is lagging behind for a no apparent reason (high chance they'll trade higher and their return will be higher than the $90 in the future) -> you should sell/trimm the position on the $90.

_________________

volatility in the portfolio is a misunderstood topic.
each indivudal is different, you could be fine with huge position in illiquid stocks/other things, e.g. BTC. while many people want to be able to liquidate at any moment. I don't think Peter Thiel, who has billions tied in illiquid startups' stock, is ever worried about the value of his investments, and he is way smarter than us and smarter than the average winning investor.

so, if you bought btc at $200 and now it's at $600 and it's up 3x and so your position is 3x bigger and it's worth more in % of your portfolio you shouldn't reduce because of volatility in value of the portfolio, more risks, etc... but, because the expected return / opportunity-costs tell you that it's smarter to sell it and invest in something else.

The potential return seems like a very weird metric to use. Let's take Bitcoin as an example. I estimate the potential return be ~$4M/BTC (2016-$s) in say a $10 years frame. I estimate the probability to be VERY low, but that is the potential. How would you then use this estimate? Care to elaborate?

I assume you don't mean to estimate the possible outcomes and assign all of them some probability and from there calculate the expected value and from that the utility for me? This just seems like a such monumental problem. Has anyone ever done this? I see an almost continuous outcome space where all the probabilities are really hard to estimate. And this for 10 different assets. Which are changing continuously.

I guess one conclusion of this discussion is that this is very hard problem and that most of us just choose some arbitrary strategy. I am fine with this. The most interesting then is just to hear everyone's strategy and then maybe we can form some heuristic strategy based on that?

I think you are on the right track heltok. You may be interested in reading about Kelly Criterion and the differences between arithmetic/geometric means and how they are used. Or anything by Aaron Brown.

I would generally avoid arbitrary heuristics that don't make any sense (Never put more than 2% of your roll on a trade for example.)

There's momentum but there's also mean reversion.

Also, you are diversified in the first place to reduce risk. Your risk tolerance may trump EV

You can often reallocate in tax deferred accounts to reduce the tax penalty.

You should probably be tax loss harvesting -- which will allow you to offset gains from reallocation.

I'm still saving so I reallocate by putting new savings into underweight buckets.

As at least one other eventually mentioned, I think kelly criterion is the relevant model to apply portfolio allocation. While the formula asks for a few inputs all of which are very non-transparent in financial markets, you can make simplifying assumptions, or work backwards, to make assessments.

Eg: "Suppose I'm 50-50 that my pick is a winner or a loser and the upside is 1.5x the downside, what allocation does that apply?"

Or, "If I have 5% of my portfolio in this trade and suppose the upside and the downside are equal, then what does the probability of going up from here have to be?" Etc.

How often you should review your allocations is: whenever things substantially change. Either the prospects for any investment, or your allocation. Which mean pretty much all the time, to varying degrees.

It's more the opposite, it's booking profits. Your guess is based on the assumption that something that's gone up is more likely to go up in future than something that's gone down.

These guys make indices based on the theory.

http://www.researchaffiliates.com/Ou...ages/home.aspx

This makes the least sense out of anything in a post generally full of investment fallacies. The price you purchased something that has no impact on it's value to you now (tax considerations aside).

are you claiming that momentum effects no longer exist in US equities? i would be interested in seeing a recent study that says this if you have one.

No, I'm not saying that. If you want to assess the value of a stock on momentum then feel free, and use that to change your strategic allocation. I was more referring to mechanistic rebalancing, i.e. how close you stay to the strategic allocation you've picked.