Gyroscopic Investing

If you were to create a 3x33% PP (i.e., cashless PP), what rebalancing bands would you use and why?

In effect I do have one because my better half is somewhat skeptical about non-cash savings and so we have >50% of our savings as cash that is not part of the same "rebalancing pool" as the 3x33% PP (ie there is never going to be rebalancing out of cash into the 3x33%). Its a compromise. If it was totally down to me, I'd have all of that cash as part of the rebalancing pool (and just have 25% cash).
What I've been doing is using 27% to 40% bands (ie 4/3x20% to 4/3x30%). My reasoning was that without the cash reserve on hand, the somewhat tighter rebalancing bands are prudent. My guess is that in choppy, to and fro markets such as we have had since 1999, such a tight banded 3x33% PP might be the best option anyway. I've never done any backtests though.

If you were to create a 3x33% PP (i.e., cashless PP), what rebalancing bands would you use and why?

 

I did start a 3x33% PP about a year ago described here:

http://gyroscopicinvesting.com/forum/ht ... ic.php?t=7

I have made some changes mostly to the individual stocks about 1 year later described here:

http://gyroscopicinvesting.com/forum/ht ... 2#msg40872

From running simulations, I did find that tighter bands do work better.  I am using 30% bands for rebalancing the 3 major components, although I have
not had a rebalance yet.  I use 40% rebalance bands for the individual stocks and had 1 rebalance of the stocks about midway into it.  I originally hadn't
considered rebalancing the stocks, but the previous responder (Stone) asked about it and suggested it.  I ran some simulations and found he was right.
Norm

If I ran a PP like this I might just rebalance annually.

If I ran a PP like this I might just rebalance annually.

I feel there are 2 things in favor of rebalancing annually:
1.  It is much easier to implement.
2.  According to the simulations I have run there is never MUCH difference in results between rebalancing methods.

However, intuitively rebalancing annually over using bands doesn't make much sense to me.  As I understand it, the whole idea of rebalancing is to sell
high and buy low thus harvesting gains.  To me, rebalancing at arbitrary time periods won't accomplish this as well as rebalancing when components are
relatively high or low (using bands).
Norm

ngcpa wrote:

If I ran a PP like this I might just rebalance annually.

I feel there are 2 things in favor of rebalancing annually:
1.  It is much easier to implement.
2.  According to the simulations I have run there is never MUCH difference in results between rebalancing methods.

However, intuitively rebalancing annually over using bands doesn't make much sense to me.  As I understand it, the whole idea of rebalancing is to sell
high and buy low thus harvesting gains.  To me, rebalancing at arbitrary time periods won't accomplish this as well as rebalancing when components are
relatively high or low (using bands).
Norm

I can confirm that my backtesting has also shown that the method of rebalancing doesn't really matter.  Some methods will slightly outperform during certain periods but will then mean revert.  I'd say just go with whichever method one is more comfortable with and stick with it.

ngcpa wrote:

If I ran a PP like this I might just rebalance annually.

I feel there are 2 things in favor of rebalancing annually:
1.  It is much easier to implement.
2.  According to the simulations I have run there is never MUCH difference in results between rebalancing methods.

However, intuitively rebalancing annually over using bands doesn't make much sense to me.  As I understand it, the whole idea of rebalancing is to sell
high and buy low thus harvesting gains.  To me, rebalancing at arbitrary time periods won't accomplish this as well as rebalancing when components are
relatively high or low (using bands).
Norm

When you look at the historical performance of the traditional PP based upon annual rebalancing, it's done pretty well.

I agree that with annual rebalancing you are potentially not capturing as much of the gains as a 15/35 rebalancing approach might provide, but you are capturing a lot of them.

PRPFX, for example, provides a very tight rebalancing band approach because of its stated allocations, and it has done very well in recent years.

The takeaway for me is that 15/35. 20/30 and annual rebalancing methodologies are all likely to generate satisfactory results, even if there is some theoretically optimal approach somewhere in that mix.

If I had to pick one I would pick 15/35 on the 4x25% approach, but with a 3x33% approach it's harder to see where the optimal bands might be.

Without cash acting as a stabilizer for the portfolio, I'd probably be more conservative with the rebalancing bands, such as 27/40. But I like round numbers which is why I chose them, since it doesn't necessarily represent a perfectly sized band (-6.33% vs +6.67%)

I had a lot more respect for 15/35 bands compared to 20/30 bands based on the simulations I ran (sorry, can't remember the difference).  I think using 20/30 bands yielded an increase in CAGR of around 0.5% compared to annual rebalancing.

If you're considering a 33x3 PP, it's because you're trying to juice returns by 1-2%, right?  So you probably won't want to pass up a 0.5% gain.

ngcpa wrote: I did start a 3x33% PP about a year ago described here:

http://gyroscopicinvesting.com/forum/ht ... ic.php?t=7

I have made some changes mostly to the individual stocks about 1 year later described here:

http://gyroscopicinvesting.com/forum/ht ... 2#msg40872

From running simulations, I did find that tighter bands do work better.  I am using 30% bands for rebalancing the 3 major components, although I have not had a rebalance yet.  I use 40% rebalance bands for the individual stocks and had 1 rebalance of the stocks about midway into it.  I originally hadn't considered rebalancing the stocks, but the previous responder (Stone) asked about it and suggested it.  I ran some simulations and found he was right.
Norm

Thanks for the links to your previous threads, Norm. By 30% rebalancing bands, I assume you mean you rebalance if any one of the three asset classes exceeds 43% (1.3*33%) or falls below 23% (0.7*33%) of the total portfolio?

sophie wrote: If you're considering a 33x3 PP, it's because you're trying to juice returns by 1-2%, right?

Actually, juicing returns isn't my main motivation for this. (Although juicy returns are always nice!) I recently conceptually restructured my assets so that instead of considering all of my taxable and retirement assets as a single PP, I've separated out a cash emergency fund, a standalone taxable PP, and a standalone retirement PP.

Since most of my taxable assets were in cash (which I converted into a 4x25% PP a couple of days ago) and I have virtually no cash in my retirement accounts, this conceptual restructuring had the net effect of turning my retirement accounts "instantly" into a 3x33% PP.

So I figure I can either start off with this 3x33% PP and just add all my new retirement contributions to cash to bring it back to 4x25% within about 3 years, or I can rebalance right now to bring everything to 4x25%. The problem with the latter approach is that it would require me to pay Schwab's dreaded $25 broker-assisted bond sale fee.

Yes, I am actually considering doing a temporary 3x33% instead of a 4x25% just to spite Schwab and their damn $25 fee. (Bastards!) The way I see it, I won't be touching those retirement accounts for a few more decades. So I think I can probably handle the increased volatility of the 3x33% for a few years while I build the cash piece back up.

Bastards, indeed.

Is the money in an IRA?  You can vote with your feet and take your business elsewhere.  Both Vanguard and Fidelity will sell bonds for you at no fee, and you should be able to transfer the assets as is.  Shame on Schwab.  Do they really believe no one will think to do this?

My Schwab account is a 401(k) at the company where I currently work, so I'm kind of stuck with it for now. About 3/4 of my retirement money is in it.

On second thought, even if I start with this 3x33% PP and gradually transform it into a 4x25% PP over 2 or 3 years, there's a good chance I'd hit a rebalancing band within that timeframe anyways--which would likely entail selling some of my bonds and paying that $25 fee I'm trying so hard to avoid. And those rebalancing bands would also have to change as the allocation gradually shifted from 3x33% to 4x25%.

Eh... maybe I'll just bite the bullet and rebalance to 4x25% tomorrow. There's a lot to be said for keeping things simple, even if it entails an occasional fee.

Why don't you call Schwab and ask them to waive the fees?  They just might if you insist that you're planning to move your money out in the near future because of that issue.

Otherwise, I guess you're kind of stuck.  It's probably not the worst thing in the world.  Maybe you can rebalance only to get 15% in cash, and then let that build up, to minimize the fees?

Norm. By 30% rebalancing bands, I assume you mean you rebalance if any one of the three asset classes exceeds 43% (1.3*33%) or falls below 23% (0.7*33%) of the total portfolio?

That is correct.

sophie wrote: Maybe you can rebalance only to get 15% in cash, and then let that build up, to minimize the fees?

I thought about doing that, but the problem is that it would result in a bond-heavy portfolio with cash at 15%, stocks and gold around 25%, and bonds at 33%. If bonds had an upward price movement of any sort, they'd likely hit the 15/35 rebalance band. By contrast, having all three volatile assets at 33% would at least keep them in balance with each other and less likely to trigger a rebalance event nearly as quickly.

sophie wrote: Why don't you call Schwab and ask them to waive the fees?  They just might if you insist that you're planning to move your money out in the near future because of that issue.

Haha... I suppose I could, but I'm not a very good bluffer. And it would definitely be a bluff since I can't move my brokerage window anywhere else at the moment. Nah, I'm leaning towards just rebalancing to 4x25%, putting a voodoo curse on Schwab because of the annoying fee, and calling it a day

Thanks for the suggestions.

sophie wrote: I had a lot more respect for 15/35 bands compared to 20/30 bands based on the simulations I ran (sorry, can't remember the difference).  I think using 20/30 bands yielded an increase in CAGR of around 0.5% compared to annual rebalancing.

If you're considering a 33x3 PP, it's because you're trying to juice returns by 1-2%, right?  So you probably won't want to pass up a 0.5% gain.

sophie, I was confused by your statement. You wrote that you have more respect for 15/35 over 20/30 bands but state that 20/30 has a 0.5% gain over annual rebalancing? I would assume if you said you had more respect for 15/35 that you'd state that had a higher gain over annual rebalancing (or a gain over 20/30 as well).