33% domestic stocks bias

masonic

itwasntme001 said:

But my point was that there is no such thing as gaining any sort of performance benefit by "rebalacning" or reducing risk when alternative assets are provding a competitive return, at least with any sort of certainty.

Suggest you revisit this old classic: https://forums.moneysavingexpert.com/discussion/5208032/the-power-of-the-rebalance

itwasntme001

masonic said:

itwasntme001 said:

But my point was that there is no such thing as gaining any sort of performance benefit by "rebalacning" or reducing risk when alternative assets are provding a competitive return, at least with any sort of certainty.

Suggest you revisit this old classic: https://forums.moneysavingexpert.com/discussion/5208032/the-power-of-the-rebalance

But that looks just at one period of 25 years, that too during a period where a lot of "mean reversion" happened, favouring a rebalancing type of strategy.  Try the same thing from 2010 till now, and you will see the 100% US fund being a lot better than rebalancing etc.

I am all for diversification, reducing risk of ruin etc, hell I reduced a large chunk of my equity allocation in my pension very recently.

But lets not pretend there is any sort of rebalancing bonus or free lunch when employing the strategy in that thread.  There isn't.

masonic

itwasntme001 said:

masonic said:

itwasntme001 said:

But my point was that there is no such thing as gaining any sort of performance benefit by "rebalacning" or reducing risk when alternative assets are provding a competitive return, at least with any sort of certainty.

Suggest you revisit this old classic: https://forums.moneysavingexpert.com/discussion/5208032/the-power-of-the-rebalance

But that looks just at one period of 25 years, that too during a period where a lot of "mean reversion" happened, favouring a rebalancing type of strategy.  Try the same thing from 2010 till now, and you will see the 100% US fund being a lot better than rebalancing etc.

I am all for diversification, reducing risk of ruin etc, hell I reduced a large chunk of my equity allocation in my pension very recently.

But lets not pretend there is any sort of rebalancing bonus or free lunch when employing the strategy in that thread.  There isn't.

I'm not pretending. Holding a basket of uncorrelated risk assets and rebalancing between them can give one an edge over allowing the same portfolio to drift freely, unless one of those assets continues to outperform consistently, which is rare in the world of risk assets.

You cannot say "there is no such thing as gaining any sort of performance benefit..." and then dismiss out of hand actual data showing this assertion to be incorrect.

If you don't like that particular example, how about the data in Bernstein's article on the subject? https://www.efficientfrontier.com/ef/996/rebal.htm

Another example is a mix of equities and gold or equities and broad commodities futures, where the rebalanced 50:50 portfolio would tend to outperform the portfolio that is allowed to drift over 40 year periods since 1970, and likely other blends of these asset classes too. Monevator had a series of articles on commodities where this was explored.

itwasntme001

masonic said:

itwasntme001 said:

masonic said:

itwasntme001 said:

But my point was that there is no such thing as gaining any sort of performance benefit by "rebalacning" or reducing risk when alternative assets are provding a competitive return, at least with any sort of certainty.

Suggest you revisit this old classic: https://forums.moneysavingexpert.com/discussion/5208032/the-power-of-the-rebalance

But that looks just at one period of 25 years, that too during a period where a lot of "mean reversion" happened, favouring a rebalancing type of strategy.  Try the same thing from 2010 till now, and you will see the 100% US fund being a lot better than rebalancing etc.

I am all for diversification, reducing risk of ruin etc, hell I reduced a large chunk of my equity allocation in my pension very recently.

But lets not pretend there is any sort of rebalancing bonus or free lunch when employing the strategy in that thread.  There isn't.

I'm not pretending. Holding a basket of uncorrelated risk assets and rebalancing between them can give one an edge over allowing the same portfolio to drift freely, unless one of those assets continues to outperform consistently, which is rare in the world of risk assets.

You cannot say "there is no such thing as gaining any sort of performance benefit..." and then dismiss out of hand actual data showing this assertion to be incorrect.

If you don't like that particular example, how about the data in Bernstein's article on the subject? https://www.efficientfrontier.com/ef/996/rebal.htm

Another example is a mix of equities and gold or equities and broad commodities futures, where the rebalanced 50:50 portfolio would tend to outperform the portfolio that is allowed to drift over 40 year periods since 1970, and likely other blends of these asset classes too. Monevator had a series of articles on commodities where this was explored.

I agree that it can (could) give one an edge over not rebalancing, but as you seem to agree as well there is no guarantee of this, hence no such thing as a free lunch in rebalancing.

When I say there is no performance benefit, you should read the full sentenace to give proper context - I said there is no performance benefit with certainty.

Your example of having commodities in the mix is also just one example (of infinately many other combinations of AA and time periods, many of these would see rebalancing as an infeiror choice), that too starting from a period which saw commodities rise and fall, zig zagging with equities bonds etc.

There is no point picking time periods to justify rebalancing bonus certainty.  At least pick a random time period and AA and do this many times.

Alexland

itwasntme001 said:
there is no guarantee of this, hence no such thing as a free lunch in rebalancing.

Since when did a free lunch need to be guaranteed?

If a free lunch happens then I don't need to have been promised it over breakfast. It's still happened.

Diversification, portfolio theory and rebalancing are known as a free lunch because they enable investors to lower risk while sacrificing little if any long term return. ie maximising risk-adjusted return.

They are certainly a lot better than investing all your money in totally the wrong thing.

itwasntme001

Alexland said:

itwasntme001 said:
there is no guarantee of this, hence no such thing as a free lunch in rebalancing.

Since when did a free lunch need to be guaranteed?

If a free lunch happens then I don't need to have been promised it over breakfast. It's still happened.

Diversification, portfolio theory and rebalancing are known as a free lunch because they enable investors to lower risk while sacrificing little if any long term return. ie maximising risk-adjusted return.

They are certainly a lot better than investing all your money in totally the wrong thing.

Free lunch by definition is something that is guaranteed, otherwse it wouldn't be a free lunch as there will always be a risk of the "free lunch" (rebalanced portfolio) being an adverse outcome to the alternative (letting portfolio drift).

I do not think any reasonable academic or practitioner would say diversifcation or rebalancing are free lunches.  They might reduce portfolio risk for not much reduction in return, but those are just expectaitons based on historical data (efficient frontier etc) and as above are by no means a certainty to happen.  Hence no free lunch.

masonic

itwasntme001 said:

masonic said:

itwasntme001 said:

masonic said:

itwasntme001 said:

But my point was that there is no such thing as gaining any sort of performance benefit by "rebalacning" or reducing risk when alternative assets are provding a competitive return, at least with any sort of certainty.

Suggest you revisit this old classic: https://forums.moneysavingexpert.com/discussion/5208032/the-power-of-the-rebalance

But that looks just at one period of 25 years, that too during a period where a lot of "mean reversion" happened, favouring a rebalancing type of strategy.  Try the same thing from 2010 till now, and you will see the 100% US fund being a lot better than rebalancing etc.

I am all for diversification, reducing risk of ruin etc, hell I reduced a large chunk of my equity allocation in my pension very recently.

But lets not pretend there is any sort of rebalancing bonus or free lunch when employing the strategy in that thread.  There isn't.

I'm not pretending. Holding a basket of uncorrelated risk assets and rebalancing between them can give one an edge over allowing the same portfolio to drift freely, unless one of those assets continues to outperform consistently, which is rare in the world of risk assets.

You cannot say "there is no such thing as gaining any sort of performance benefit..." and then dismiss out of hand actual data showing this assertion to be incorrect.

If you don't like that particular example, how about the data in Bernstein's article on the subject? https://www.efficientfrontier.com/ef/996/rebal.htm

Another example is a mix of equities and gold or equities and broad commodities futures, where the rebalanced 50:50 portfolio would tend to outperform the portfolio that is allowed to drift over 40 year periods since 1970, and likely other blends of these asset classes too. Monevator had a series of articles on commodities where this was explored.

I agree that it can (could) give one an edge over not rebalancing, but as you seem to agree as well there is no guarantee of this, hence no such thing as a free lunch in rebalancing.

I agree with the bit in bold, but not the bit in italics. A "free lunch" is getting something for nothing. There is no requirement for the free lunch to happen in all circumstances, only some. When a rebalancing bonus happens, it is a free lunch. Diversification is also widely described as a free lunch, because it can allow investors to reduce risk without sacrificing potential returns, but that too doesn't always work.

itwasntme001 said:

When I say there is no performance benefit, you should read the full sentenace to give proper context - I said there is no performance benefit with certainty.

The full sentence was (my emphasis): "But my point was that there is no such thing as gaining any sort of performance benefit by "rebalacning" or reducing risk when alternative assets are provding a competitive return, at least with any sort of certainty."

This is a rather more absolutist statement, with the "at least" acting as a safety net, and the use of the emphatic negation at the end widens the net to suggest even the most basic level of confidence could not be reached. Essentially, this read to me as if you were saying there is no such thing, and even if there was it could almost never be predicted.

That's where I am disagreeing.

itwasntme001 said:

Your example of having commodities in the mix is also just one example (of infinately many other combinations of AA and time periods, many of these would see rebalancing as an infeiror choice), that too starting from a period which saw commodities rise and fall, zig zagging with equities bonds etc.

There is no point picking time periods to justify rebalancing bonus certainty.  At least pick a random time period and AA and do this many times.

It seems you don't like any of my examples. It probably won't help to give you more, as I've not the time or the data to hand to do a full systematic analysis. So I think we'll just have to agree to disagree on the probability of being able to harvest a rebalancing bonus.

itwasntme001

masonic - it appears our debate (and with Alexland) comes down to our respective definitions of what a free lunch is.

We agree that at any given moment, deciding to rebalance vs drift a portfolio is by no means a guarantee that the former will perform better than the latter.

You say "probability of. ..... rebalancing bonus", which I agree that this probability increases and decreases as markets moves around over time - e.g. this probability imo has increased for a simple portfolio of equities and bond compared to 4 years ago.

But for something to be a free lunch, there needs to be no cost at all under all scenarios.  There might certainly be a cost, even with high probaility of a rebalancing bonus, for employing a strategy of rebalancing vs just drift, if indeed the drifting portfolio produces a better return.

masonic

I can't help but feel that this extreme definition of "free lunch" makes the term meaningless. Which is fine, because I didn't introduce it into the discussion and have no issue with dispensing with it, but I don't think anything in the real world has no cost at all under all scenarios. I suppose you could counter with, hence TANSTAAFL.

itwasntme001

masonic said:

I can't help but feel that this extreme definition of "free lunch" makes the term meaningless. Which is fine, because I didn't introduce it into the discussion and have no issue with dispensing with it, but I don't think anything in the real world has no cost at all under all scenarios. I suppose you could counter with, hence TANSTAAFL.

Do you not think that some of your comments or Alexland's and that Bowlhead thread you posted, might be misconstrued to believe rebalancing is actually a free lunch with a guaranteed bonus by some people reading it?

Hell, I suspect that a whole generation of academics working on portfolio theory, efficient frontier, rebalancing in the late 20th century might have resulted in a persistant risk premium of sorts in equities, which has in more recent years unravelled to produce above average returns.  At what cost to long term patient investors?  Or indeed benefit for more newer investors?