Rebalancing questions

[nic]

I have some questions about rebalancing:

1. How are the PP rebalancing percentages obtained?

2. How important are the percentages -- i.e. how much could you suffer by modifying them? I have in mind here rebalancing at lower levels, say 10% or so.

3. Is there any reason to use unsymmetric rebalancing with different percentages up and down?

3. Is there any data to show what would happen (in theory) if you rebalance continuously in time?

4. More generally, is there any data to show what happens if you rebalance discretely in time instead of in asset value?

Thanks for your help with these questions.

Nic

[LNGTERMER]

Clive, if possible can you expand on this ladder method.

I also came to the conclusion that in general the MA system does not work. It looks great on charts but over a long run it's mediocre where zig-zags and fees eat up most of the gains in addition to considerable lag built in which gives back the gains. So, it ends up to a system that locks you into losses but gives a big margin of your profits, hence does not work.

[LNGTERMER]

Clive, much thanks for taking the time and explaining, I am sure this will be of use for a lot of others as well.
I must admit, this is something that never occurred to me so, it might take me a while to wrap my head around it.
Nevertheless, this can be a system I might use for my VP portion of the folio. Thanks again.

[LNGTERMER]

If I understand this correctly, it seems you divide the asset/stock into some discrete steps while defining the top and bottom, then you keep adding when a step is breached and reducing by same amount when the step is recovered. Am I correct here?

Assuming I am, then in worst case you can end up in a situation where you will never recover your money. For example if you chose QQQQ in year 2000 at a price of 114 with sufficiently large step and deep bottom, then even thought we have been zigzagging all along we are still locked in and potentially will never recover that money in our life time. Am I missing something?

[LNGTERMER]

Clive, thanks for the valuable feedback.
I guess, when looked at for a specific asset in isolation it seems a bit risky, however, if one uses the pairing you mentioned it might tilt the odds in one's favor.

The AIM methodology, particularly seems interesting, I find the pyramidinvesting.com description a bit unclear. I tried to research the AIM, but the site simply is not revealing much, it seems they are selling it through books and such.

[foglifter]

I'd like to continue this thread and ask our rebalancing experts what they would recommend to those who tend to be more comfortable with a simpler rebalancing strategy (if one is possible at all).

1. Are there any grounds behind HB's recommended 15/35 rebalancing bands or are they purely theoretical?

2. Do you know of any statistics comparing performance of PP with different rebalancing bands?

2. Is it feasible to do some sort of "tactical rebalancing" - that rebalancing based on specific asset's condition (pre-bubble, steady growth etc.)

[LNGTERMER]

Foglifter, I ran many rebalancing scenarios. My conclusion is that it works and that it is a corner stone of the strength of the PP along with its dis-correlated asset mix. I found very little difference between the bands. My ranges went from 5 to 35%. They all worked and that reduced the choice to a personal one.
I did not run the 15 down band verses the 35 up bands, so I can let you when I get the chance. I will post the actual system output later.

[craigr]

The rebalancing bands Harry Browne stated were not based on any scientific calculation. Rather they just seemed like a good number that has worked through the years for him. I tend to agree.

My rebalancing tends to happen between the 30-35% range on the high side depending on what's needed for tax reasons and harvesting gains vs. losses in the year. Profits are always used to rebalance the lowest performer back up first and the remainder used to buy the second loser, etc. I never allow anything to go above 35% and never below 15%. Most of my rebalancing seems to happen near the 35% end of the scale. I try to really not do much to the portfolio and prefer avoiding the gains as long as possible to lower tax and transaction costs.

However the above really is if you have a specific tax planning need. If you don't have that, then I would stick to a hard number that you decide (30%, 32.5%, whatever) and don't complicate things.

As for statistics, I've seen some people post the numbers. I've run them myself several years ago but unfortunately lost the spreadsheet somewhere along the way. My take away is that "it depends" (as usual). Sometimes it is an advantage to wait to the 35% range to capture the momentum, but sometimes you wait too long and the price never hits the 35% range and goes back down. It's very time dependent. I would pick a number and stick to it or, like I said above, if you have some specific tax planning need that requires you to sell before the upper 35% band that is probably OK, too. There is an argument that the 30% upper band could reduce volatility in the portfolio for the more squeamish.

I wouldn't mess with any kind of tactical rebalancing. This really is trying to guess what the market is going to do and isn't a good idea.

[LNGTERMER]

I second that, but just to make sure we are talking about the same thing, when I say 5-35% I am talking about an asset gaining that percentage. If, however, you mean an asset weight becoming 35% instead of the original 25% then that also is within the range of my simulations. I would recommend staying with HBs' recommendation, it's simple and works.

[foglifter]

I second that, but just to make sure we are talking about the same thing, when I say 5-35% I am talking about an asset gaining that percentage. 

If I'm reading your numbers correctly it means the asset share in the portfolio can rise to anywhere from 30% to 60%! Nothing wrong with that - you might capture the upside gains and I did see this in my pre-PP portfolio. As craigr mentioned the problem is to catch the right time to rebalance so you don't lose that gain if the asset starts slipping back.

[LNGTERMER]

That is right, I was testing for extreme cases. I personally settled in rebalancing every time an asset increases/looses it weight by at least 5% of total PP. These translates to 30% and 15% bands. The 10% are less work and they work just as well.

[foglifter]

That is right, I was testing for extreme cases. I personally settled in rebalancing every time an asset increases/looses it weight by at least 5% of total PP. These translates to 30% and 15% bands. The 10% are less work and they work just as well.

Interesting - so it means asymmetric bands: fluctuations from -10% to +5%.

[LNGTERMER]

I'll try to be alittle more clearer  . The asset starts with a weight of 25% then when it goes up to 35% you rebalance. If it goes down to 15% then also rebalance. My personal choice is up band of 30% or down band of 20%, but there is not a very meaningful difference between the two though.

[foglifter]

I'll try to be alittle more clearer  . The asset starts with a weight of 25% then when it goes up to 35% you rebalance. If it goes down to 15% then also rebalance. My personal choice is up band of 30% or down band of 20%, but there is not a very meaningful difference between the two though.

OK, so you must have mistyped 15% instead of 20% and that's what confused me.