Canadian Money Forum banner

241 - 260 of 274 Posts

·
Registered
Joined
·
15,884 Posts
Discussion Starter #241
I see what you mean. Yes I can see how it all depends on how much fixed income and how much withdrawal is happening.

I would want to try simulating through some bear market scenarios before I'm convinced on which strategy is best.
 

·
Registered
Joined
·
532 Posts
Also some posts have commented on not rebalancing. If that is in response to my post #231 above, my suggestion was to keep rebalancing into fixed income during bull markets, but not rebalance into equities during bear markets.
That would be a divergence from the "asset allocation" portfolios. In classical asset allocation, rebalancing is done on schedule (or based on bands), from stocks to bonds and vice versa without exception. If one only rebalances from risk assets to safe assets, the portfolio will have different characteristics. The closest that I could think of is McClung's Prime Harvesting, which only rebalances one way.

I think what AltaRed does is having a good safe reserve that is only deployed when needed. That probably will be closer to the bucket strategy that has been described by Christine Benz et al. So for example at retirement one puts 1m into dividend paying stocks that pay say 50k in dividends per year. If that 50k covers the needs of the retiree, he will make a determination that a cash or safe bond cushion of 250k is needed to cover the shortfall of a 50% dividend cut for 10 years. There is no rebalancing unless the reserves get substantially depleted. In essence it is an 80/20 portfolio with no rebalancing, with a reasonabe SWR of 4%. If equities do well and the dividends double to 100k, theoretically there is no need for the cash cushion anymore.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #243
That would be a divergence from the "asset allocation" portfolios.
I think so too.

Keep in mind, there's also the "rising equity glidepath" idea. The idea there is: the most dangerous times for a retiree are the early years, the first ~ 10 years which includes the lead-up to retirement.

At that early point it makes sense to be very heavy in fixed income. So one could start retirement with a high % fixed income allocation. However once you're past that 10 year danger period, you can safely start increasing equity allocations and can even keep increasing equities into old age.

Perhaps that's something to consider for rebalancing. In the early stages of retirement one can be very conservative and only rebalance equity-to-bonds but not the reverse direction (which will increase % bonds). However once you're past the first decade, then you could resume regular rebalancing.

But again I'd want to simulate to be sure. Of course IMO the easiest answer is simply to increase your fixed income allocation when in doubt. Especially today, after we'd had such spectacular equity performance, there is nothing lost by increasing fixed income/GICs.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #244
I'll describe my own approach to this. I'm living off capital, but not quite retired.

I find this idea of maintaining multiple buckets rather awkward. Instead, I've rolled all my fixed income needs into a big fixed income allocation (50% FI). I know others have other allocations like 70/30 and then do a separate mental accounting for things like GICs, which in reality means they are nowhere close to 70% stocks. My reasoning is: keep it simple and roll it all into your FI allocation.

I think this also helps me be more honest about my overall risk position. It's harder to know where you stand if you have 70% stocks in one place, but extra fixed income buckets elsewhere. What's the net positioning after all that?

Within that fixed income allocation, I have a mix of liquid bonds (XBB and government) and many GICs. Since there's a healthy amount of illiquid GICs, there's no danger that I will rebalance during a bear market and deplete my fixed income. What it means in practice is that the liquid side gets used for rebalancing whereas the GIC ladder is there forever. Plenty to live off, if I ever need to.
 

·
Registered
Joined
·
9,957 Posts
Keep in mind, there's also the "rising equity glidepath" idea. The idea there is: the most dangerous times for a retiree are the early years, the first ~ 10 years which includes the lead-up to retirement.

At that early point it makes sense to be very heavy in fixed income. So one could start retirement with a high % fixed income allocation. However once you're past that 10 year danger period, you can safely start increasing equity allocations and can even keep increasing equities into old age.
Intuitively, this is about where I was at retirement in 2006. About 40% fixed income and it coincidentally served me well in 08/09 to be at circa that number. More than 13 years later, I am now about 15% fixed income BUT that is an outcome (consequence) of what I have chosen to hold in my 'cash equivalents' reserve, not because I want it to be 15%. If equity markets go up significantly again in 2020, then my fixed income percentage would drop.

That all said, managing my fixed income cushion is an ongoing dynamic process depending on what near term plans may be, e.g. a $35k holiday or vehicle replacement. As an example, I will most likely be buying a new vehicle in 2020 or 2021. As soon as 2020 arrives, I will be selling an equity at substantial capital gains as part of the plan to build that reserve in preparation for that transaction whenever it may occur.
 

·
Registered
Joined
·
532 Posts
I'll describe my own approach to this. I'm living off capital, but not quite retired.

I find this idea of maintaining multiple buckets rather awkward. Instead, I've rolled all my fixed income needs into a big fixed income allocation (50% FI). I know others have other allocations like 70/30 and then do a separate mental accounting for things like GICs, which in reality means they are nowhere close to 70% stocks. My reasoning is: keep it simple and roll it all into your FI allocation.

I think this also helps me be more honest about my overall risk position. It's harder to know where you stand if you have 70% stocks in one place, but extra fixed income buckets elsewhere. What's the net positioning after all that?

Within that fixed income allocation, I have a mix of liquid bonds (XBB and government) and many GICs. Since there's a healthy amount of illiquid GICs, there's no danger that I will rebalance during a bear market and deplete my fixed income. What it means in practice is that the liquid side gets used for rebalancing whereas the GIC ladder is there forever. Plenty to live off, if I ever need to.
I think that is close to an asset allocation portfolio, although your GICs could be considered a separate bucket/emergency fund. In most studies on SWRs using asset allocation portfolios there is no limit on how much can be rebalanced.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #247
I think that is close to an asset allocation portfolio, although your GICs could be considered a separate bucket/emergency fund. In most studies on SWRs using asset allocation portfolios there is no limit on how much can be rebalanced.
Right, and it has occurred to me that I might not have enough bond liquidity for some really huge rebalancing. I want to calculate what the most extreme rebalancing transfer would be from bonds to stocks, say in a Great Depression kind of scenario. Here's a quick calculation covering 1929-1932 which was an extremely bad period

Starting with 500k each in stocks and bonds (50/50 allocation) in 1929 and assuming no withdrawals,

end of 1929: sell 31.2k in bonds (6% of bond portfolio), buy stocks
end of 1930: sell 72.6k in bonds (14% of bond portfolio), buy stocks
end of 1931: sell 90.7k in bonds (21% of bond portfolio), buy stocks
end of 1932: sell 29.4k in bonds (8% of bond portfolio), buy stocks

Totals: 223.9k of bonds sold, 223.9k of stocks bought. How's that for "buy low"?

Look at that insanity! Would anyone do this in real life? Sell a cumulative 223.9k in bonds out of an original 500k while the stock market crashes over 4 years? That's what's required for rebalancing.

That 1930 year also requires 21% of the bond portfolio to be sold. Again, would anyone do that? Liquidate the one asset which just saved your neck, pour the money into a black hole?

I think this illustrates how difficult rebalancing can be in a big stock market crash. If you did this, the result is of course excellent because you've bought a cumulative 223.9k in stocks during the crash... but can anyone do that in real life?

Seeing this result makes me feel better about my GICs. I am happy limiting the rebalancing by holding illiquid, super safe fixed income.
 

·
Registered
Joined
·
9,957 Posts
One might do some of that if in a secure job accumulating assets, but few were feeling secure during that period. Like everything else, a little common sense needs to be applied to conventional wisdom.
 

·
Registered
Joined
·
532 Posts
It is very difficult and that is why one needs to be committed and do this very mechanically. Let's not forget that the market crashed again a few years later.

One could also rebalance once or twice and then just watch and wait so as to preserve capital.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #250
One might do some of that if in a secure job accumulating assets, but few were feeling secure during that period. Like everything else, a little common sense needs to be applied to conventional wisdom.
Right. For example I can imagine that a government worker, or a wealthy young person with a trust fund, or perhaps someone with a solid pension could still rebalance the ideal way during that stock crash.

I'm also sure some pensions were less "solid" after 1929.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #251
It is very difficult and that is why one needs to be committed and do this very mechanically. Let's not forget that the market crashed again a few years later.

One could also rebalance once or twice and then just watch and wait so as to preserve capital.
Yup. In fact a more thorough back test is needed because as you say, the stock market remained weak for a very long time... this wasn't just a 4 year thing.

I still have no idea how I would react mysef, to rebalancing within wild movements. All the basic asset allocation stuff we talk about is great over a period like 1980-2019 but I'm not so sure about other, worse periods.
 

·
Registered
Joined
·
904 Posts
I still don't quite understand this part. Is it to make sure you still have lots of fixed income, so you don't risk depleting it during a long bear market?
Yes, you do understand. You explained it in post #247. It is to avoid the situation you described here:

Right, and it has occurred to me that I might not have enough bond liquidity for some really huge rebalancing. I want to calculate what the most extreme rebalancing transfer would be from bonds to stocks, say in a Great Depression kind of scenario. Here's a quick calculation covering 1929-1932 which was an extremely bad period

Starting with 500k each in stocks and bonds (50/50 allocation) in 1929 and assuming no withdrawals

end of 1929: sell 31.2k in bonds (6% of bond portfolio), buy stocks
end of 1930: sell 72.6k in bonds (14% of bond portfolio), buy stocks
end of 1931: sell 90.7k in bonds (21% of bond portfolio), buy stocks
end of 1932: sell 29.4k in bonds (8% of bond portfolio), buy stocks

Totals: 223.9k of bonds sold, 223.9k of stocks bought. How's that for "buy low"?

Look at that insanity! Would anyone do this in real life? Sell a cumulative 223.9k in bonds out of an original 500k while the stock market crashes over 4 years? That's what's required for rebalancing.

That 1930 year also requires 21% of the bond portfolio to be sold. Again, would anyone do that? Liquidate the one asset which just saved your neck, pour the money into a black hole?

I think this illustrates how difficult rebalancing can be in a big stock market crash. If you did this, the result is of course excellent because you've bought a cumulative 223.9k in stocks during the crash... but can anyone do that in real life?

Seeing this result makes me feel better about my GICs. I am happy limiting the rebalancing by holding illiquid, super safe fixed income.
In most market conditions, if you never rebalance, equities will grow faster, resulting in your asset allocation being heavier in equities than you want, and your portfolio volatility and risk will gradually increase. So over a long enough time you will need to eventually rebalance from equities to FI if you want to maintain your target asset allocation.

But in a long, deep bear market, equities will crash but should eventually recover and continue up. So in a bear market you don't need to sell fixed income and buy equities to maintain your asset allocation. Just sit tight and time will rebalance automatically for you.

It avoids the gut clench you would get in times like 2008, 2000, 1987, 1973, or worst case 1929 when you would watch stocks continue to fall by 85% over 4 years. As you said "Look at that insanity! Would anyone do this in real life? " Instead you just sit tight and wait for stocks to recover on their own. It also avoids the apocalypse situation that would happen if you sold bonds to buy stocks that never recover (say for exaple they are a big Ponzi scheme).
 

·
Registered
Joined
·
2,963 Posts
In classical asset allocation, .
I don't know what Classical asset allocation is.

I do recall financial advisers and pundits recommending that investors hold 100-age in equity. This link suggests even 110 or 120 - age in equity. So the experts have changed their allocation recommendations in a relatively short time frame. Why? Can't they make up their minds? If we had followed that advice, portfolio would have yielded much less and we would have had to sell off assets to cover living expenses. In other words, portfolio would have depleted and we would be worse off.

It just makes no sense to me to look at allocation on a percentage of portfolio basis.

As an example, when we retired in 2003, we had 43% of portfolio in FI.
12 months later, the percentage had dropped to 38%. (dollar amount in FI was unchanged.)
At end of 2008, FI was 55% of our portfolio. (dollar amount in FI down a bit)
Now FI is just 35% (dollar amount in FI higher)

I have only increased FI dollar amount marginally. (@ a bit less than inflation rate). The amount is our safety cushion and is, with CPP/OAS, more than enough to carry us through for a decade or more.

My advice - Just put enough in a safe place to live on frugally for protracted period. Add a little from time to time as living costs go up. If still working, aim at a FI number that you would like at retirement and build your portfolio that way.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #255
Yes, you do understand. You explained it in post #247. It is to avoid the situation you described here:
...

In most market conditions, if you never rebalance, equities will grow faster, resulting in your asset allocation being heavier in equities than you want, and your portfolio volatility and risk will gradually increase. So over a long enough time you will need to eventually rebalance from equities to FI if you want to maintain your target asset allocation.

But in a long, deep bear market, equities will crash but should eventually recover and continue up. So in a bear market you don't need to sell fixed income and buy equities to maintain your asset allocation. Just sit tight and time will rebalance automatically for you.

It avoids the gut clench you would get in times like 2008, 2000, 1987, 1973, or worst case 1929 when you would watch stocks continue to fall by 85% over 4 years. As you said "Look at that insanity! Would anyone do this in real life? " Instead you just sit tight and wait for stocks to recover on their own. It also avoids the apocalypse situation that would happen if you sold bonds to buy stocks that never recover (say for exaple they are a big Ponzi scheme).
Thanks. You make a very strong argument here.
 

·
Registered
Joined
·
10,362 Posts
... I'm also sure some pensions were less "solid" after 1929.
It would take some digging to get a full picture but a new points from US based articles are that the number of pension plans *grew* from 1929 to 1932, by fifteen percent. In contrast, three percent are reported as the number of pensions that were discontinued.

Keep in mind that the pensions are described as being a lot different than the more recent ones. To apply one had to have twenty years of service, be aged sixty and be recommended by one's manager. If the board of directors plus the pension committee approved the application then one would get half of one's salary to a maximum of five hundred dollars a year.

Lots of room for pension costs to be dealt with by rejecting newer pension applications, with AFACIT no federal gov't pension legislation to protect the worker.


The first US pension is reported to have been setup in 1875, where until 1921, pension contributions were taxed by the IRS.

It seems to be about fifteen percent of the US workforce that had pensions in the early '20's, though it is unclear if this includes both public and private pensions. It wasn't until the 1940's that labour became interested in pensions where it took until the 1950's for twenty five percent of the private sector workforce to have a pension.


It looks like the US gov't passed pension legislation as a response to failing pensions in 1974 so there's lots of possibilities for how the '20's pensions were funded.


Cheers
 

·
Registered
Joined
·
2,963 Posts
It is what is used in studies to determine the sustainable withdrawal rate, for example a 60/40 portfolio that rebalances once a year in December.
To be honest, I don't see what was a "Classical Allocation" in those studies.

I read that they chose a series of arbitrary Equity/FI ratios from 0-100% and back tested them from 1925 to 1995 and later to 2009 to determine what withdrawal rate could be sustained. I haven't read this in detail recently, but ISTR that they figured 50-75% stocks at start of retirement was optimal.

OK, here is the original report: http://www.retailinvestor.org/pdf/Bengen1.pdf

As already pointed out. I think those FI/Eq ratios are of little value anyway, except in accumulation stage as a guide.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #258
I see more value than you're suggesting, agent99.

First of all, those studies have also been done with monte carlo simulations. These are more powerful than simple back tests because they simulate potential paths (the range of possible outcomes) based on statistical properties of stocks & bonds. It has repeatedly been shown that asset allocations within a certain range ... generally anywhere from 30% to 70% equities ... provide the best outcome for withdrawals.

These are not arbitrary ratios. The researchers have figured out what mixes work well. For example something like 50/50 or 60/40 is not accidental and isn't pulled out of thin air. It's believed to provide quite good outcomes in most possible stock & bond future scenarios.

And all of these assume the classic AA, meaning rebalancing to hold a steady allocation.
 

·
Registered
Joined
·
15,884 Posts
Discussion Starter #260
The Bengen study is the gold standard for retirement SWRs. If we change one variable (such as the frequency of rebalancing), we have to be cognizant of the fact that the SWR could change too.
Right. These are the parameters of the problem as it has been studied so far.

This isn't to say that other methods cannot work. Maybe different rebalancing strategies, or living off dividends, are completely viable alternate solutions. However, these studied did not consider them.

Similarly, the kind of things I do (permanent portfolio / risk parity) have not been thorough studied either for withdrawal and capital depletion. I can't say conclusively that it's a good idea, or that it's any better than classic asset allocation for living off capital.

And if I was a professional, I would absolutely not be able to advise anyone to use these methods on just a hunch and my amateur analysis.
 
241 - 260 of 274 Posts
Top