Gyroscopic Investing

I fully agree that the quantitative analysis may be a bit much, but I did it for one main reason:  I don't like doing something arbitrary without understanding why I'm doing it.  So, my initial question was "why 15/35?"

From the analysis I did, it does turn out that the exact rebalancing band doesn't matter as much as I thought it might.  Rebalancing every day is within a percentage point of rebalancing at a 5/45 band!  That's pretty interesting information right there, if you ask me.  I'm sticking with 15/35, myself.  My curiosity has been satisfied.

What are current thoughts regarding this? And, more importantly, what IS YOUR rebalancing method, e.g., how frequently, what triggers it?

Vinny

Rebalancing benefits or costs are completely path dependent. There is really no solid "evidence" statistically for one method over another. That's the bottom line. The standard line, which I concur with, is that rebalancing is primarily about risk not performance.

In/assuming a taxable account, one can mess around with different methods to see what the tax hit is which is a useful exercise. Such an approach looks at frequency of trades and assesses the distribution of ST and LT capital gains taxes and then adjusts performance expectations accordingly.

Monthly deposits go into cash. I check the portfolio annually with a calendar reminder, and also after a major finance-related world event that everyone's talking about (e.g. 9/11 or the 2008 global financial crisis). If it's within the 15/35 bands I do nothing, otherwise I rebalance to 4x25. (IIRC this is precisely Browne's advice on the radio show.)

For the first 10 years or so of accumulation, after every annual check cash was overweight and I bought the other assets. Now that the portfolio is larger relative to deposits, I don't necessarily have to rebalance every time. So far I've never sold stocks/bonds/gold.

Did these past week's events cause any of you to rebalance? If so, what bands that you are using were crossed?

Vinny

Think it's a bit early to rebalance yet. A 10% drop in one asset, especially one that was probably over-weighted going in, isn't going to shift your asset percentages by more than 3-5%.

However, this is absolutely a tax loss harvest opportunity! I was going to do it on Friday, but decided to wait and see what happened today. So far, that was the right move. Anyone thinking same?

I've never tax-loss harvested before, but understand the basic concept. You can't purchase the same security for 30 days though to avoid the wash sale rule though, right? So what are you thinking of selling, and will you simply sit in cash for the 30 days, potentially missing a bounce back as central banks flood the world with liquidity?

drumminj wrote: ↑Mon Mar 02, 2020 9:11 am I've never tax-loss harvested before, but understand the basic concept. You can't purchase the same security for 30 days though to avoid the wash sale rule though, right? So what are you thinking of selling, and will you simply sit in cash for the 30 days, potentially missing a bounce back as central banks flood the world with liquidity?

My understanding is that there isn't a formal definition of when two investments are "too close" and thus fall under wash sale rules.

If you're swapping an S&P 500 index for a total market index, I think it would be hard for the IRS to claim those are the same. Or a closed-end gold fund for an open-ended gold fund. Or a Treasury with 20 years left for one with 30 years left.

Basically, I think (but nobody can say for sure) that there are a lot of ways to get sufficient coverage during the 30-day window without falling afoul of the wash sale rules.

I harvested some IJS (S&P 600 value) for VBR (CSRP small cap value).

Also, I put my full yearly bonus that I got on Friday into stocks, which is technically a rebalance without selling anything. It worked out quite nicely as for the last 4 months or so while stocks had been blowing up all my new cash was going into bonds during their pullback. Now bonds are ripping, stocks are pulled back, and my fresh funds are all going into stocks. I like buying long term assets when they are down, I hate having to buy an asset that is at a high, though I still do it if the asset is the lowest.

I spoke too soon yesterday :-) but looks like the market will drop again today, so I'm pulling the trigger.

I switch to a stock fund just different enough to avoid the wash sale. For Vanguard I switch between the passive and tax managed fund versions. For Fidelity it's a bit harder. I have to switch between total market and small cap funds. You hold the new fund for at least 30 days. If it gains a lot in the meantime, I'll just hang on to it. If it loses, all the better, just sell it to claim some more losses.

WARNING: check to make sure you haven't done any reinvestment of dividends/gains in the last 30 days!!! And obviously don't sell any lots less than 30 days old.

KevinW wrote: ↑Tue Feb 04, 2020 12:36 pm Monthly deposits go into cash. I check the portfolio annually with a calendar reminder, and also after a major finance-related world event that everyone's talking about (e.g. 9/11 or the 2008 global financial crisis). If it's within the 15/35 bands I do nothing, otherwise I rebalance to 4x25. (IIRC this is precisely Browne's advice on the radio show.)

For the first 10 years or so of accumulation, after every annual check cash was overweight and I bought the other assets. Now that the portfolio is larger relative to deposits, I don't necessarily have to rebalance every time. So far I've never sold stocks/bonds/gold.

So....how many times have you checked in 2020??!!

Vinny

moda0306 wrote: ↑Sat Oct 08, 2011 10:46 am
I think it's safe to say the 15/35 bands work better historically because all three volatile assets had long periods of ascent and decline.  This is bound to result in wider bands being beneficial.  I'm not saying things won't continue that way, but if we have smaller, choppier swings out of these assets in the coming years the smaller bands will work better going forward.

Now with the benefit of being able to add in ten more years to "historically"...you still hold to the same analysis?

edsanville wrote: ↑Sun Oct 09, 2011 8:48 pm
I've run a few backtests using different rebalancing bands.  The tests only go from 2004-present, because I used ETF data to do them, (GLD, TLT, VTI).  Here's what I found:

Rebalancing band: 25%-25%   Annualized returns: 8.770407%  (rebalance every day no matter what!)
Rebalancing band: 24%-26%   Annualized returns: 8.736676%
Rebalancing band: 23%-27%   Annualized returns: 8.730892%
Rebalancing band: 22%-28%   Annualized returns: 8.633693%
Rebalancing band: 21%-29%   Annualized returns: 8.609356%
Rebalancing band: 20%-30%   Annualized returns: 8.801298%
Rebalancing band: 19%-31%   Annualized returns: 8.746533%
Rebalancing band: 18%-32%   Annualized returns: 8.755651%
Rebalancing band: 17%-33%   Annualized returns: 8.593875%
Rebalancing band: 16%-34%   Annualized returns: 9.144876%
Rebalancing band: 15%-35%   Annualized returns: 9.225508%
Rebalancing band: 14%-36%   Annualized returns: 9.533811%
Rebalancing band: 13%-37%   Annualized returns: 8.762806%
Rebalancing band: 12%-38%   Annualized returns: 8.644704%
Rebalancing band: 11%-39%   Annualized returns: 8.620474%
Rebalancing band: 10%-40%   Annualized returns: 9.013725%
Rebalancing band:  9%-41%   Annualized returns: 9.110751%
Rebalancing band:  8%-42%   Annualized returns: 8.773225%
Rebalancing band:  7%-43%   Annualized returns: 8.848743%
Rebalancing band:  6%-44%   Annualized returns: 8.962410%
Rebalancing band:  5%-45%   Annualized returns: 9.100235%

So, it looks like the results agree fairly well with the other posters here:  rebalancing more often doesn't improve your yields.  The 15-35% band seems to be in a "sweet spot,"  although that could just be a lack of data.  If you take into account transaction costs, of course, then the more frequent balancing will lose a bit more.

Have any of you done a more recent, updated similar analysis?

moda0306 wrote: ↑Mon Oct 10, 2011 12:10 pm
I really think using the past rebalance comparisons to give any kind of false precision on rebalance bands is a bit much.

The longer an asset tends to trend, year by year, to do the same thing as it did in prior years, the more you will benefit from wider bands.

If you end up with a relatively long period of all three assets swinging wildly in opposition to their prior-year behavior, a 30/20 band setup will work better.

Lastly, assuming the correlations hold up, if you are at 34/16 of two of the assets, you are in a position where one is over twice as heavily weighed as the other.  This could really hammer you if the higher asset retreats strongly, and with only 16% of a diversifier to help soften the blow.

This isn't to say I don't believe in wider bands.  In fact, I think I'd stick with 35/15 bands.  I just like to fully understand the reason wider bands have succeeded, and the risks I'm running by assuming it will behave that way in the future.

The thing I like about 35/15 bands, is that they are a very good way, over a lifetime, of simultaneously capturing momentum (waiting longer to sell a booming asset), and using some of that gain to buy other assets low, which allows the portfolio to become more than a sum of its parts.

In fact, I actually thing, especially for a taxable account rebalance, doing partial rebalances back to 20/30 from 35/15 is appropriate.  It keeps them in the "risk zone" you want them in, further rewards your best assets by on paring them back so much, and is more tax-efficient.

Agreement / disagreement on the above highlighted point?

Just going to go in here without reading the whole thread. Listened to the Harry Browne radio show a while ago and it was surprising to hear how non quantitative his approach to rebalancing was.

He said rebalance when one of the assets gets too big. Basically imagine a 50 percent drawdown in then asset class and assess how heavily it would impact your overall portfolio.

I’ll throw up the link to the show if I can find it.

For PP rebalancing, I think the theme has been discussed to nauseating extremes Though I cannot really recall of the Golden Butterfly rebalancing being discussed somewhere (nor in Tyler's site), any idea ?

PS. He (Tyler) recently published a new article on his site, apparently he is going through some personal stress and health issues. Hope he is doing all fine.

You are right about Tyler. I think he hurt his leg or knee on an electric scooter. I am going to reach out and wish him well.

christina wrote: ↑Sat Oct 08, 2011 9:15 am I'm a little bit confused about when you're supposed to rebalance.
are you supposed to monitor your portfolio constantly, and rebalance immediately when one investment goes beyond one of the bands?
or are you supposed to pick a date, once a year, to look at your portfolio, and then rebalance at the bands? (I think Harry b. recommends this.)
or are you supposed to use your judgment about when to rebalance, as I belive craigr uses? (I believe I would not be able to do this successfully. I think I would prefer a more mechanical approach.)

what about rebalancing quarterly? I heard that returns might be better with more frequent rebalancing, but I'm not sure. I will likely be holding my portfolio as a set of ETFs that are freely tradable, so trading costs are not an issue for me.

Thanks!

Rebalancing is generally considered as being a risk-reduction practice, where otherwise a target asset allocation might have drifted considerably into being a entirely different asset allocation.

For retirees no rebalancing other than taking income from the most-up asset might suffice. Asset drift tends to be less of a issue with time, as typically higher volatility relative to already secured good gains is less of a risk (if you see a 33% portfolio drop after having achieved a 50% gain you've only just given back 'other peoples money').

Not rebalancing, letting winners run, often leads to higher rewards than rebalancing.

You can use PV to inspect historic differences, click the Allocation Drift, change the start dates and change whether rebalanced or not ...etc.

I'm inclined to opine that for younger/saving drip feeding into stocks is more inclined to accumulate the larger retirement pot. When transitioning from accumulation into drawdown early years sequence of returns risk is the greater risk period and as such shifting into a PP is a reasonable means to lower that risk. But then just leave that as-is, no rebalancing, which will tend to see higher rewards as that generally shifts more weight into stocks over time, excepting if a high early years sequence of returns risk does come to light, in which case shifting cash/bonds/gold into stocks is reasonable, but very discretionary. All-stock even in retirement years can be fine provided the stock wasn't purchased at too high a price. All-stock is suggested as having a 4% historic SWR i.e. worst case, if the stock were purchased at a 33% discount then that sees the SWR rise to 6%.