Personal Finance

TD e-series rebalancing

  • Last Updated:
  • Apr 3rd, 2010 9:24 pm
Deal Addict
User avatar
Oct 2, 2005
1801 posts
1446 upvotes
Montreal
WalnutCrunch wrote: Wow. What an extreme example! :D

Rebalancing may or may not increase returns, but one important thing it does do is return our portfolios to their set asset allocations, therefore returning our investments to our predefined risk profile.
Rebalancing to get back to your initial stock / bond allocation is fine.
The problem I have is with rebalancing country specific funds, which people tend to underestimate in terms of just how risky they are.

A study showed that investing in 1 country specific index fund for 10 years is just as risky as investing in a global index fund for 1 year, in terms of probability of loss. A specific country can consistently under-perform for a decade, even a century. In my example above, by rebalancing every year, I was actually taking more risk by moving most of my money into 1 country.
Deal Addict
Feb 20, 2006
1186 posts
48 upvotes
Vancouver
No you were taking less risk because your actual capital was more balanced across countries. It's not how much you move where that matters for future-looking risk, but how much you actually end up with allocated in each place.

Or are claiming that leaving it alone, and have $900 in Bangladesh and $50 in the US is less risky?
Deal Addict
User avatar
Oct 2, 2005
1801 posts
1446 upvotes
Montreal
Sanchez wrote: No you were taking less risk because your actual capital was mire balanced across countries. It's not how much you move where that matters for future-looking risk, but how much you actually end up with allocated in each place.

Or are claiming that leaving it alone, and have $900 in Bangladesh and $50 in the US is less risky?
I would take the extra profits from Bangladesh and put it in something else; I wouldn't keep pouring it into the US fund, trying to maintain 25% allocation.

If you had a portfolio of Canadian stocks 10 years ago, cap weighted to replicate the market, you would've started with a 33% allocation to Nortel. Should you have kept pouring money into Nortel over the years to maintain your initial asset allocation?
Deal Addict
Feb 20, 2006
1186 posts
48 upvotes
Vancouver
Noobzilla wrote: I would take the extra profits from Bangladesh and put it in something else; I wouldn't keep pouring it into the US fund, trying to maintain 25% allocation.
So you admit you would rebalance, but are now using the benefit of hindsight to claim that you would have chosen some other sector or asset class that you knew would rise over the next 10 years? It's not exactly clear then what you methodology/rule is at all then - just invest in the things that are going to do the best over the next few years? Not very implementable...
If you had a portfolio of Canadian stocks 10 years ago, cap weighted to replicate the market, you would've started with a 33% allocation to Nortel. Should you have kept pouring money into Nortel over the years to maintain your initial asset allocation?
Huh? No, as Nortel's stock price and thus market cap dropped you would automatically remain correctly allocated to Nortel (and all other equities) in proportion to it's market cap - no re-balancing is necessary.
Deal Addict
User avatar
Oct 2, 2005
1801 posts
1446 upvotes
Montreal
Sanchez wrote: So you admit you would rebalance, but are now using the benefit of hindsight to claim that you would have chosen some other sector or asset class that you knew would rise over the next 10 years? It's not exactly clear then what you methodology/rule is at all then - just invest in the things that are going to do the best over the next few years? Not very implementable...
If you keep pouring new money into 1 fund, for whatever reason, that's not risk reduction, that's an increase in risk - you started off diversified, but now you're allocating new capital in a non diversified manner.

Think of what an investment advisor in Japan might have recommended for a couch potato strategy in 1989:

40% Nikkei index
40% international index
20% bonds

Rebalance once a year. Great, you started off with a well balanced portfolio.
As the Nikkei continually falls, you would put your paycheck towards the Nikkei index every month get it back up to 40%. This new money is being applied in a non diversified manner, and this strategy will ensure that if you have 1 badly performing fund, it will drag down the rest of your portfolio with it. Many Japanese investors did this, lost nearly everything, and completely gave up on investing. The average household in Japan now has only 3% of it's net worth allocated to equities, versus about 30% for US households.

Now, I didn't need to make any predictions about the Nikkei to see that this was a risky strategy. If the investor would have ensured that new money was being applied in a diversified manner, the risk of lost would have been reduced. Put your next paycheck towards a Brazilian small cap ETF; anything but the Nikkei every time.

Sanchez wrote: Huh? No, as Nortel's stock price and thus market cap dropped you would automatically remain correctly allocated to Nortel (and all other equities) in proportion to it's market cap - no re-balancing is necessary.
Right, so there's no need to keep pouring money into Nortel. If the US crumbles like Japan did, there's no need to keep pouring money into your US fund to get it back up to 25% of your portfolio. But that's what our coach potato strategy would tell us to do whenever it's time to rebalance or add new money. That's why I said - be careful with rebalancing; make sure it's not going into the same fund every time.
Deal Addict
Feb 20, 2006
1186 posts
48 upvotes
Vancouver
Noobzilla wrote: If you keep pouring new money into 1 fund, for whatever reason, that's not risk reduction, that's an increase in risk - you started off diversified, but now you're allocating new capital in a non diversified manner.
No, risk is about how your current capital is allocated, not how all your new capital has been allocated over your investing lifetime.

Furthermore, you can't predict in advance whether re-balancing is "pouring all your money into 1 fund" or smartly buying low in advance of that fund rebounding.
Think of what an investment advisor in Japan might have recommended for a couch potato strategy in 1989:

40% Nikkei index
40% international index
20% bonds

Rebalance once a year. Great, you started off with a well balanced portfolio.
Yes, anyone who invested in heavily Japan didn't do too well over the past 20 years. However, I believe that someone with that portfolio would still have done fine, even with re-balancing, although not re-balancing certainly would have been the better strategy in this specific case.

The problem with any investment rule is that you'll be able to find a scenario where another rule will work far better. For example, choose an obviously insane rule like "invest all your money in stocks at a 2:1 leverage ratio". This rule will actually work better than any more conservative strategy 4 times out of 5. Does that make it better? Of course not, because the other 20% of the time you got wiped out. So your Japanese example hardly proves the point.

Furthermore, you give no clue on what your rule actually is. You say "you shouldn't have rebalanced into the Nikkei over the past 20 years" - but everyone knows that now! A better rule, if you are going to use hindsight, is simply not invest in the Nikkei at all, not to mention re-balancing.

How did you know that at the time? Plenty of other markets have dropped, and then come back strongly. You would generally have been a fool not to re-balance into most market drops in, across most markets in history. How do you pick out the few Nikkei's which did occur (and that story isn't over)? One answer is you don't - you re-balance mechanically and take the emotion out of it. Another answer is that you cap-weight your asset allocation across countries, just like you do within a country. In this case there is no need to re-allocate equity from one country to another when one country drops, although a re-allocation from other asset classes (e.g., FI) would occur.
Now, I didn't need to make any predictions about the Nikkei to see that this was a risky strategy. If the investor would have ensured that new money was being applied in a diversified manner, the risk of lost would have been reduced. Put your next paycheck towards a Brazilian small cap ETF; anything but the Nikkei every time.
Huh, of course you had to make a prediction to see that this was risky - because you picked the worst period for a highly unusually behaving market. Your example fails most of the time. Furthermore, your example is silly because re-balancing is not just about reducing risk - but keeping risk in a pre-defined interval. This means that about half the time, it is all about increasing risk, since your less risky assets have become over-weighted. The dropping Nikkei is one of those cases - rebalancing here will increase risk, because the portfolio has been over-weighted in less risky assets.

In the reverse situation, where equity outperforms FI, do you argue against "pouring money into bonds"? Do you allow the equity portion of your portfolio to approach 100% of your AA simply because you don't want to pour money into an asset which hasn't performed? This is a very risky strategy.



Right, so there's no need to keep pouring money into Nortel.
Did you read what I wrote in my prior reply? Any cap weighted strategy (the traditionally recommended ones) wouldn't have poured an additional dime into Nortel since its cap weight is dropping exactly in proportion to its stock price. Re-balancing for a cap weighted indexer should primarily between asset classes, not individual stocks.
If the US crumbles like Japan did, there's no need to keep pouring money into your US fund to get it back up to 25% of your portfolio. But that's what our coach potato strategy would tell us to do whenever it's time to rebalance or add new money. That's why I said - be careful with rebalancing; make sure it's not going into the same fund every time.
What exactly is your rule. Don't use words like "crumbles" because you have no idea if a drop is the beginning of a crumble, or a small dip representing a buying opportunity. You've got to implement your re-balancing strategy without knowing what the markets will actually do, so what is your rule? Do you have a threshold, like "after a market has dropped xy% from its peak, stop adding any money to it"?
Deal Addict
User avatar
Oct 2, 2005
1801 posts
1446 upvotes
Montreal
Sanchez wrote: No, risk is about how your current capital is allocated, not how all your new capital has been allocated over your investing lifetime.
The global index will change over your lifetime. Empires fall, as is the case with every empire in history, and new ones rise. If you keep trying get your portfolio back to how things used to be when you started, you will deviate from the world index as it changes, leading you to be overweight in certain sectors. Hence, you will have a riskier portfolio than a global index.

The most passive investing approach is to buy a cap weighed global index fund. Got more to invest? Add more to the fund.

Let’s say you start with:
60% global index
40% bonds

If your international index goes up to 78% allocation and you want to get back to your intended level of risk, rebalancing makes sense. Stocks, due to their nature, are clearly riskier.

But what if you started with this:
25% Canadian index
25% US index
25% international index
25% bonds

Let’s say the US index goes down another 20% this decade and another 20% the next decade, while the others do fine. You would keep adding most new funds to the US index for decades without giving up? What was the point? Were you taking less risk than simply adding new funds to a global index? No. Did you feel that that lower prices were a good buying opportunity? Maybe they are, but why limit yourself to 1 country? Why not buy low according to something more meaningful, like the P/E ratio or price/book ratio? If you have new money to invest, and want the best value for your money, you can go to Yahoo Finance and sort through 900 different ETFs according to P/E ratio, 1 year return - anything you want. Surely there are other good opportunities besides the US.

I can understand maintaining 25% allocation to the Canadian index so that currency risk doesn’t get out of hand.
Sanchez wrote: You say "you shouldn't have rebalanced into the Nikkei over the past 20 years" ... How did you know that at the time?
For one thing, the P/E ratio of the Nikkei was 78. The higher the P/E ratio, the less earnings you get for your money. It doesn’t indicate much about what will happen next year, but is says a lot about what’s likely to happen to your investment over the next 20 years.

[IMG]http://upload.wikimedia.org/wikipedia/c ... ata%29.png[/IMG]

If a country index started with a 10 year trailing P/E ratio of 78, and the index dropped 80% over the next 20 years, does that make it a good buying opportunity now? Not if the P/E ratio is still high. There’s no reason to expect Japan to now outperform any other country.

If a country index had a P/E ratio of 20 and quickly lost 50% in 3 months, so that it now has a P/E ratio of 10, is it a good buying opportunity? Possibly; things tend to get oversold during a panic.

Rebalancing can help you avoid loses from a speculative bubble, unless you started during a bubble.
Deal Addict
Feb 20, 2006
1186 posts
48 upvotes
Vancouver
You make a good point about re-balancing between counties. I mentioned that above - you may not need to re-balance at all between countries or regions if you are following a cap-weighted strategy, because losses will be directly proportional to reduced market cap, so your weighting will remain correct without any need to re-balance - just as in the Nortel example.

If you aren't following a market cap strategy (e.g., because of home country bias) then things get a bit more complicated, as you point out with your 25% in Canada example.

Re-balancing mostly applies between distinct asset classes (bonds, other forms of FI, equity), etc.

Top

Thread Information

There is currently 1 user viewing this thread. (0 members and 1 guest)