Rebalancing

[sixdollars]

Hello all.  I've frequented this forum (silently) for a while now, but never got around to actually posting any comments or topics. I'm hoping to bounce some ideas off the community here.  I was originally a Bogleheads style investor (mostly just stock index funds / bonds) but recently switched to the PP.  Just some background, what led me to the PP was the realization that economic growth (my opinion) can't go on forever.  For a while, I blindly believed that continued economic growth was a good assumption and this assumption allowed me to feel comfortable with a heavy stock allocation as the stock market would "always go up".  Somewhere along the way, I came to realize there are always physical limits, and with the limitations of a finite planet, I now see my previous assumption was a bit foolish, and so I set out to search for a better investment method.  The PP allows me to sleep much more comfortably now  . 

Anyway, I have been thinking for a while about PP rebalancing and one thing has me a little bit concerned.  I was hoping to bounce a question off in this forum to see what everyone's thoughts are on long term rebalancing.

Can anyone think of any situations where constantly rebalancing would actually cause you to lose all/most of your wealth?  Are there any scenarios you can think of where one or two of the PP assets would continually decline and where constant rebalancing from the winning asset to the losing asset would cause your wealth to approach zero?  In most potential scenarios, I imagine that this would happen with Gold being the winning asset but one of the other assets in continual decline (maybe stocks?).  I realize that this question would only be applicable to a black swan type of event, but I'm still curious as to whether or not it will always be prudent to rebalance agnostically when the bands are reached.  Just trying to see if there's a hole to be poked in the rebalancing method.

Under most circumstances, it does seem like a good idea.. but would there be things (signals?) you would look out for in the future that might indicate to you that conventional rebalancing techniques will no longer work going forward?  Are there certain situations where other measures should be taken to protect your wealth?  Any contribution to the discussion is much appreciated.

[AdamA]

http://gyroscopicinvesting.com/forum/pe ... /#msg20397

Here's a link from a few years ago where this was nicely discussed.

Craig made the point that the PP is about having options.  It gives you more of chance to reevaluate the investing environment and react appropriately than any other strategy that I'm aware of.  (It's explained more articulately in the above link).

[Jan Van]

For an alternative view on rebalancing:

Did You Know That Asset Allocation and Rebalancing Don’t Work?

[Pointedstick]

For an alternative view on rebalancing:

Did You Know That Asset Allocation and Rebalancing Don’t Work?

Logically, rebalancing should result in a penalty for an asset that tends to go up forever. You'll be constantly selling to maintain a percentage, gradually reducing the number of shares. Taking profits to rebalance is pointless when waiting will simply result in more profits. Similarly, with an asset that goes down forever, you'll be continuously piling into a sinking ship. The "buying low" aspect of rebalancing is transformed into a vice because it is always going lower, causing you to spend more and more money on it. No good.

However, for a set of assets that tend to go up and down over different time periods, and that move opposite one another at various times, none ever going to zero, we should expect rebalancing to not have either of those drawbacks--in fact, it should force us to buy low and sell high.

Ryan Melvey explored this two years ago using the example of a coin toss: http://www.stableinvesting.com/2013/04/ ... demon.html

Basically, it seems like the existence of a "rebalancing bonus" is heavily dependent on your choice of things to rebalance.

[buddtholomew]

I choose not to rebalance by selling out-performing assets. I prefer to buy lagging assets in an attempt to restore the HBPP allocation to within 15/35 bands. Not sure where this approach falls in the "to rebalance or not" discussions. Also, I choose to rebalance incrementally (2-3% at a time), in an attempt to mitigate the impact of an ever-declining asset.

[coinstar]

Since I'm in the accumulating phase, I rebalance by buying lagging assets with new money. There's only two real strategies that make any sense for new money:

1) Put it all in cash until cash hits a 35% rebalancing band, then do a rebalance on all assets
2) Buy lagging assets with new money to maintain 25%x4 at all times (or close to it, maybe I go between 23% and 27%)

I suppose there COULD be a third way to "rebalance" with new money, and that's to put new money in such that it's a 4-way split. Basically putting the new money in as it's own mini-PP, equal parts of each, regardless of what ratios the portfolio you have exists at.

And MAYBE a FOURTH way: add new money to maintain whatever the current split is. So if you're 20% gold, 22% bonds, 30% cash, 28% stocks, then all of the new money gets added in that same ratio. That would keep you from buying too much of the losers and keep you buying MORE of the winners to get momentum effect.

However, those 3rd and 4th methods would have way too much overhead. Suppose I get $2k to invest this month. Splitting that up 4 ways would mean I can't actually buy a $1k face-value t-bill/t-bond, nor can I buy a 1 ounce gold coin. From a logistics perspective, just putting that entire $2k into the lagging most asset is most efficient. (assuming your portfolio is $100k+ and that $2k is only a small percentage of the whole. If you have a $10k portfolio and get $2k new money, then all in one asset probably won't work)

[Tyler]

Basically, it seems like the existence of a "rebalancing bonus" is heavily dependent on your choice of things to rebalance.

+100

Rebalancing in and of itself is not magical.  Rebalancing between highly correlated assets (like some advice to slice stocks 20 different ways) generally just eats up fees with little benefit.  Rebalancing between a handful of volatile uncorrelated assets addressing different market conditions is another thing entirely. 

[sophie]

However, for a set of assets that tend to go up and down over different time periods, and that move opposite one another at various times, none ever going to zero, we should expect rebalancing to not have either of those drawbacks--in fact, it should force us to buy low and sell high.

Ryan Melvey explored this two years ago using the example of a coin toss: http://www.stableinvesting.com/2013/04/ ... demon.html

Basically, it seems like the existence of a "rebalancing bonus" is heavily dependent on your choice of things to rebalance.

Rebalancing is part of the PP's "secret sauce".  The four PP assets were chosen not only as an elemental representation of the entire market, but because they can often be anti-correlated - not just "not" correlated.  For example, we are constantly told that gold has zero real yield.  That may be true, but when it moves opposite to bonds & stocks it can provide big rebalancing bonuses.

Check out the great discussion in this thread:  http://gyroscopicinvesting.com/forum/pe ... arvesting/

[sixdollars]

Craig made the point that the PP is about having options.  It gives you more of chance to reevaluate the investing environment and react appropriately than any other strategy that I'm aware of.  (It's explained more articulately in the above link).

That is definitely what I love most about the PP - having options. 

Rebalancing is part of the PP's "secret sauce".  The four PP assets were chosen not only as an elemental representation of the entire market, but because they can often be anti-correlated - not just "not" correlated.  For example, we are constantly told that gold has zero real yield.  That may be true, but when it moves opposite to bonds & stocks it can provide big rebalancing bonuses.

Right, overall I agree with you.  I think the beauty of the negative correlation you describe is summarized well by William Bernstein (http://www.efficientfrontier.com/ef/0adhoc/harry.htm).  The PP seems to behave in a way that can allow for big rebalancing bonuses; however, my concern is not with this behavior in ordinary situations.  Right now I'm actually concerned about what rebalancing could do to your total weath in an extreme scenario... say one in which the price of gold might skyrocket while other assets might be on the decline.

MediumTex used an obvious example to illustrate where you might reconsider the normal rebalancing method; however, my concern is whether or not it will always be this obvious going forward.  How can you tell when a decline is going to be a transient event versus a steady state event?  I guess my biggest concern was the fact that most times it's already going to be hard to rebalance into your losing asset, yet most of us here will rebalance anyway when the bands are hit.  What's to say that going forward we won't grit our teeth and do the same - essentially rebalancing into the black hole.

I know none of us can predict the future... I just thought it might be a fun thought experiment to see if there were ways to modify the rebalancing strategy to make it more resilient.  It was touched upon that the Variable Portfolio could be used to insure against this going forward.  Solutions like this intrigue me

[rickb]

I know none of us can predict the future... I just thought it might be a fun thought experiment to see if there were ways to modify the rebalancing strategy to make it more resilient.  It was touched upon that the Variable Portfolio could be used to insure against this going forward.  Solutions like this intrigue me

You could split your assets into a "normal PP" and a "momentum rebalanced" PP (actually VP) following a modified Mebane Faber (Ivy Portfolio) rebalancing discipline.  Faber's approach is you invest or switch into cash (per asset) at the end of each month depending on the 10 month simple monthly moving average (Current price above the SMA?  Stay in or get in.  Price below the SMA?  Stay out or get out.)  You could combine this with rebalancing to determine when to buy/sell each asset.  Below your rebalance band but price below the SMA?  Don't rebalance this asset yet (wait a month and see if the price is above the SMA).  Above your rebalance band but price above the SMA?  Also don't rebalance this asset yet (wait a month and see if the price is below the SMA).  This would let winners run (without rebalancing) until the current price drops below the SMA.  This would also keep you from rebalancing into an asset that is in a prolonged decline. 

I'm not recommending this (it likely leads to extremely unbalanced allocations, which increases your risk of a significant loss due to a sudden decline in a single asset), and haven't done any backtesting (so it might prove to have been a terrible idea in some actual period in the not too distant past) - but it seems like it might address the specific situation you're concerned about. 

Another variation might be "stop loss rebalancing".  Instead of selling at peak minus 10% (or whatever your stop loss is), rebalance only after exceeding a rebalance band (on the high side) and then declining by 10%.  Flipping this around on the low side would mean rebalancing after dropping below a rebalance band only when 10% higher than the most recent low.

You could do more complicated versions of each of these by adding absolute rebalance bands as well, i.e. do momentum or stop-loss rebalancing but if an asset reaches your absolute band (maybe 5%/45%) rebalance anyway.

If anyone's interested enough in any of these to do some backtesting, let us know what you find.  My guess is which one of these three approaches wins (normal rebalancing, momentum rebalancing, stop-loss rebalancing) depends entirely on the specific data series that you test them against.  Staying roughly balanced all the time (the normal approach) is definitely the least risky.

Missing only a few of the best and worst days of the stock market (and I assume this pertains to gold and LT bonds as well) dramatically affects returns.  See https://www.ifa.com//12steps/step4/miss ... worst_days .  The real problem with any kind of momentum or stop-loss approach is that although you might miss some significant down days you're likely to miss the significant up days as well - in fact, since these signals are always late I actually suspect you're likely to have to generally endure the worst down days but generally miss the best up days.

[MachineGhost]

Can anyone think of any situations where constantly rebalancing would actually cause you to lose all/most of your wealth?  Are there any scenarios you can think of where one or two of the PP assets would continually decline and where constant rebalancing from the winning asset to the losing asset would cause your wealth to approach zero?  In most potential scenarios, I imagine that this would happen with Gold being the winning asset but one of the other assets in continual decline (maybe stocks?).  I realize that this question would only be applicable to a black swan type of event, but I'm still curious as to whether or not it will always be prudent to rebalance agnostically when the bands are reached.  Just trying to see if there's a hole to be poked in the rebalancing method.

That could to happen to three assets under The Greater Depression.  But it would only be in nominal terms up to that point confidence is restored in the cash currency or a new one world currency.

[MachineGhost]

How can you tell when a decline is going to be a transient event versus a steady state event?  I guess my biggest concern was the fact that most times it's already going to be hard to rebalance into your losing asset, yet most of us here will rebalance anyway when the bands are hit.  What's to say that going forward we won't grit our teeth and do the same - essentially rebalancing into the black hole.

Well all investments are political, short of gold, so pay attention to whats happening to global macroeconomics, such as Europe with its impending collapse of the Eurodollar or EU, what may you.

[Reub]

Thanks, Martin, er, I mean MG!