Canadian Money Forum banner

281 - 300 of 317 Posts

·
Registered
Joined
·
16,333 Posts
Discussion Starter #281
You can find a thorough study of PP together with other popular portfolios at Permanent Portfolio. According to the author of this website, the SWR(safe withdraw rate) is as high as 5.3% for 30-yrs retirement.
librahall, I'm happy to hear you found this thread (and concept) interesting.

Interesting source for the SWR. I agree that it looks like PP is able to sustain higher withdrawals due to the added stability. However I think we should consider that gold has only traded freely since the 1970s, so there is limited pricing history for it. There is a much longer history of data for stocks and bonds, so SWR on things like 50/50 or 60/40 can probably be stated with higher certainty.

Since the historical SWR analysis with gold is only based on 48 years of gold prices, there may not be enough history to be very confident.
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #283 (Edited)
Could you please comment on the following implementation of the PP in Canada? I have done some comparison and backtest in the past months. It seems US PP has outperformed CA PP for >1% CGAR in the past 20 years.
There will certainly be differences in outcomes due to performance of assets in the specific countries (this will always fluctuate over time) and even the currency.

XUS.TO (CAD unhedged) 25%
ZTL.F (CAD hedged) 25%
ZFS.TO and GIC/CD ladder 25%
MNT.TO and Physical Gold 25%
Some comments and first impressions. XUS is a good one for S&P 500 but I don't think it's a great idea to concentrate all your stock exposure into a single country. To some degree, this is chasing the recent performance of the S&P 500. It's probably a better idea to diversify the stocks somewhat. In mine, I mix Canada & US. Another method would be to split between Canada & XAW to get world exposure. At first glance, yes, this would drop the performance because the S&P 500 has performed the best, but you have to think of what might happen going forward.

And you really need some domestic stocks for the PP. The idea after all is to bring together assets which respond differently, counter-balancing economic conditions in YOUR country. If you mix up countries and currencies then you are losing that effect.

Next comment is on bonds. I don't think ZTL is a great idea, because (1) it's brand new with no track record (2) you should probably stick with bonds in your home country and (3) currency hedging has typically added performance drag on other ETFs. I think going with "TLT" or equivalent is also a form of chasing the recent performance of the US. American long term treasuries happen to have done well, but that doesn't make them inherently a better PP choice than, say, Canadian treasuries.

So I would be more inclined to go with 25% ZFL and that fund is very respectable; $1.6 billion assets under management and about 10 years of history.

I like MNT for gold but in recent times I have started diversifying some of that into CGL.C which serves the same purpose. I split mine between MNT & CGL.C for safety, because gold-tracking funds are a very new invention and I don't have full confidence in any of them. The bid/ask spreads on MNT have also been extremely wide lately, which makes it tricky to trade.

But overall, if I may suggest a slightly different take on what you posted, perhaps something like:

Stocks: 12.5% XUS, 12.5% XIC
Long bonds: 25% ZFL
Short bonds: 25% ZFS and GIC/CD ladder
Gold: 25% MNT, CGL.C, physical

1) Currency hedge. After reading many articles, I tend to believe that I should not hedge S&P 500 ETF but should hedge US Long Term Bonds. Do you think so and why?
It's true that foreign stocks should be unhedged. But the idea of holding foreign bonds and hedging them is quite a new concept, and I am not sold on the idea. As stated above, I think the PP idea would be to hold domestic (Canadian) fixed income. Consider for example, if there is runaway inflation in the US, but not in Canada. If you held this 25% TLT (US), currency hedged, you would see your "long bonds" part of PP crash in value as TLT would crash as interest rates soar in the US. Now, if you were in the US, you would have also seen your GOLD gain in value versus the USD. But notice that in your portfolio construction, you would suffer the US treasuries price crash without the balancing effect from gold, because if the same high inflation isn't happening in Canada, "gold in CAD" would not soar.

In my view, all the assets should be domestic in CAD. The only exception I have made in mine is some foreign stock component.

2) ZTL.F is newly released by BMO in Feb. this year, right at the start of the Covid-19 market crash. It's the first CAD hedged TLT in Canada.
As mentioned above I think going the ZFL & ZFS route is superior.
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #284
But does it make sense to be long cash and short a mortgage?
Well by this logic, anyone with a mortgage shouldn't hold a 60/40 balanced fund either. They are both long and short the same fixed income market.

I'm sure many balanced fund investors do have mortgages, though. I would argue that it's probably just best to aggressively pay off all debts, including mortgages, before investing too heavily into anything.
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #285
I'll also add the comment that one of the challenges with the PP (and it's entirely psychological) is that it does hold some very volatile asset classes. Gold and long term treasuries will be very volatile over the years, and swing around by huge amounts.

So although the PP overall behaves very nicely, those segments like long term treasuries will swing around a lot. You have to remember to do annual rebalancing and also stick with your portfolio. For example if interest rates go up significantly, ZFL would plummet in value. As it should. At that point you will likely feel the temptation to get out of ZFL and instead go into another bond fund, maybe entirely into ZFS. You have to withstand that temptation and stick with your annual rebalancing, meaning you would sell other things and buy ZFL and get back to your target % allocations.

That kind of thing could be extremely difficult to do, when the time comes. It's the same problem in 50/50 and 60/40, but I think it's actually a bit more difficult with PP because both long term bonds and gold are quite volatile.
 

·
Registered
Joined
·
17 Posts
But does it make sense to be long cash and short a mortgage?
librahall, I'm happy to hear you found this thread (and concept) interesting.

Interesting source for the SWR. I agree that it looks like PP is able to sustain higher withdrawals due to the added stability. However I think we should consider that gold has only traded freely since the 1970s, so there is limited pricing history for it. There is a much longer history of data for stocks and bonds, so SWR on things like 50/50 or 60/40 can probably be stated with higher certainty.

Since the historical SWR analysis with gold is only based on 48 years of gold prices, there may not be enough history to be very confident.
But gold as a good method of preserving value has been there for a few thousand years, longer than any fiat money.
 

·
Registered
Joined
·
17 Posts
But does it make sense to be long cash and short a mortgage?
I won't view holding cash as a pure way of "long". You can find the following explanation in the book of <The Permanent Portfolio> by Craig Rowland; J. M. Lawson.

"Most financial advisors recommend that investors keep some cash reserves on hand, but almost no investment strategies work these cash holdings into an overall investment strategy. However, the Permanent Portfolio is unique in that it calls for a 25 percent allocation to cash and these cash holdings are a fundamental building block of the overall strategy. Unlike the other Permanent Portfolio assets, which are designed to be volatile, cash is designed to act as a stabilizer to the portfolio during market volatility. The cash allocation also provides an investor with a place to store interest, dividends, and capital gains from the other assets, and provides “dry powder” for rebalancing purposes during market declines. Cash also occasionally serves as the leading asset in the portfolio when the stocks, bonds, and gold are all having a bad year. In addition to the functions described above, cash also acts as an emergency reserve for life's unexpected events, such as periods of unemployment or health emergencies. For those in retirement, the cash portion of the portfolio can be tapped for living expenses. Overall, the cash allocation in the Permanent Portfolio performs several important functions."
 

·
Registered
Joined
·
17 Posts
There will certainly be differences in outcomes due to performance of assets in the specific countries (this will always fluctuate over time) and even the currency.



Some comments and first impressions. XUS is a good one for S&P 500 but I don't think it's a great idea to concentrate all your stock exposure into a single country. To some degree, this is chasing the recent performance of the S&P 500. It's probably a better idea to diversify the stocks somewhat. In mine, I mix Canada & US. Another method would be to split between Canada & XAW to get world exposure. At first glance, yes, this would drop the performance because the S&P 500 has performed the best, but you have to think of what might happen going forward.

And you really need some domestic stocks for the PP. The idea after all is to bring together assets which respond differently, counter-balancing economic conditions in YOUR country. If you mix up countries and currencies then you are losing that effect.

Next comment is on bonds. I don't think ZTL is a great idea, because (1) it's brand new with no track record (2) you should probably stick with bonds in your home country and (3) currency hedging has typically added performance drag on other ETFs. I think going with "TLT" or equivalent is also a form of chasing the recent performance of the US. American long term treasuries happen to have done well, but that doesn't make them inherently a better PP choice than, say, Canadian treasuries.

So I would be more inclined to go with 25% ZFL and that fund is very respectable; $1.6 billion assets under management and about 10 years of history.

I like MNT for gold but in recent times I have started diversifying some of that into CGL.C which serves the same purpose. I split mine between MNT & CGL.C for safety, because gold-tracking funds are a very new invention and I don't have full confidence in any of them. The bid/ask spreads on MNT have also been extremely wide lately, which makes it tricky to trade.

But overall, if I may suggest a slightly different take on what you posted, perhaps something like:

Stocks: 12.5% XUS, 12.5% XIC
Long bonds: 25% ZFL
Short bonds: 25% ZFS and GIC/CD ladder
Gold: 25% MNT, CGL.C, physical



It's true that foreign stocks should be unhedged. But the idea of holding foreign bonds and hedging them is quite a new concept, and I am not sold on the idea. As stated above, I think the PP idea would be to hold domestic (Canadian) fixed income. Consider for example, if there is runaway inflation in the US, but not in Canada. If you held this 25% TLT (US), currency hedged, you would see your "long bonds" part of PP crash in value as TLT would crash as interest rates soar in the US. Now, if you were in the US, you would have also seen your GOLD gain in value versus the USD. But notice that in your portfolio construction, you would suffer the US treasuries price crash without the balancing effect from gold, because if the same high inflation isn't happening in Canada, "gold in CAD" would not soar.

In my view, all the assets should be domestic in CAD. The only exception I have made in mine is some foreign stock component.



As mentioned above I think going the ZFL & ZFS route is superior.
First of all, thank you for writing back to me during the weekend. Happy Easter Day!

The reasons why I concentrate on the US stocks are:
1) US stock market accounts for 42% of the global market in size while TSX only accounts for 2.5%. Owning 50% CAD stocks seems a home-country bias IMO;
2) US stock market has a good intrinsic diversification across different industries and even different countries already;
  • Many U.S. corporations have a global presence, with assets and revenues in foreign markets. ... At the country level, nearly 71% of S&P 500 revenue comes from the U.S., with the remaining 29% coming from foreign markets.
  • S&P 500 contains companies from 11 different sectors, while TSX 60 is weighted heavily on Finance and Natural Resource.
3) US stock market outperforms other markets with a reason, because of its overall strength in technologies, innovations, attractions to talents, capital and so on. I think this momentum may not change in my lifetime. Of course, there will be fluctuations and value recorrections along the way.

To your comments on bonds, I agree that ZTL may not be a good choice. However, I am still not quite convinced that Canadian bonds are superior to TLT(US) to hedge the US stock market. At least no historical data support this. Maybe because investors tend to buy US treasures as safe haven during the US market crash, or because US treasures have a better structure? I don't know. I need more time to think it through.

Thank you again.
 

·
Registered
Joined
·
12,184 Posts
I won't view holding cash as a pure way of "long". You can find the following explanation in the book of <The Permanent Portfolio> by Craig Rowland; J. M. Lawson.

"Most financial advisors recommend that investors keep some cash reserves on hand, but almost no investment strategies work these cash holdings into an overall investment strategy. However, the Permanent Portfolio is unique in that it calls for a 25 percent allocation to cash and these cash holdings are a fundamental building block of the overall strategy. Unlike the other Permanent Portfolio assets, which are designed to be volatile, cash is designed to act as a stabilizer to the portfolio during market volatility. The cash allocation also provides an investor with a place to store interest, dividends, and capital gains from the other assets, and provides “dry powder” for rebalancing purposes during market declines. Cash also occasionally serves as the leading asset in the portfolio when the stocks, bonds, and gold are all having a bad year. In addition to the functions described above, cash also acts as an emergency reserve for life's unexpected events, such as periods of unemployment or health emergencies. For those in retirement, the cash portion of the portfolio can be tapped for living expenses. Overall, the cash allocation in the Permanent Portfolio performs several important functions."
So, if doing PP while having debt, maybe it makes sense to hold gold and equities along with capacity to borrow for rebalancing purposes. So a notional cash/fixed income position.
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #290 (Edited)
To your comments on bonds, I agree that ZTL may not be a good choice. However, I am still not quite convinced that Canadian bonds are superior to TLT(US) to hedge the US stock market. At least no historical data support this. Maybe because investors tend to buy US treasures as safe haven during the US market crash, or because US treasures have a better structure? I don't know. I need more time to think it through.
Happy Easter to you as well

You're right that TLT is the best hedge for the US stock market, but remember, the permanent portfolio's goal is to hedge more broadly than this. The intention is to create a hedge/balance between different economic conditions including currency devaluation. It's not just about the stock/bond pairing. For example there is also an important bond/gold pairing.

So yes I agree that TLT(US) and XUS go together just fine, but, there are other assets in the portfolio that need to be appropriately matched too.

Imagine that while the CAD is strong (low inflation), the US experiences high inflation, the USD weakens and treasuries fall sharply. Here's a possible response:

XUS falls due to USD falling [stocks down or weak]
TLT falls in reaction to high US inflation [bonds down]
Gold falls, since CAD is strong [gold down or flat]

In this scenario, the PP wouldn't work as desired because you've lost the ideal inflation or currency devaluation hedge. In particular, look at the bonds & gold relationship. Here, your US bonds suffered due to inflation but gold will not compensate.

To get the ideal response on the bond/gold pairing, they have to be against the same currency, which is why I'm arguing that the PP works best with domestic assets.
 

·
Registered
Joined
·
17 Posts
Happy Easter to you as well

You're right that TLT is the best hedge for the US stock market, but remember, the permanent portfolio's goal is to hedge more broadly than this. The intention is to create a hedge/balance between different economic conditions including currency devaluation. It's not just about the stock/bond pairing. For example there is also an important bond/gold pairing.

So yes I agree that TLT(US) and XUS go together just fine, but, there are other assets in the portfolio that need to be appropriately matched too.

Imagine that while the CAD is strong (low inflation), the US experiences high inflation, the USD weakens and treasuries fall sharply. Here's a possible response:

XUS falls due to USD falling [stocks down or weak]
TLT falls in reaction to high US inflation [bonds down]
Gold falls, since CAD is strong [gold down or flat]

In this scenario, the PP wouldn't work as desired because you've lost the ideal inflation or currency devaluation hedge. In particular, look at the bonds & gold relationship. Here, your US bonds suffered due to inflation but gold will not compensate.

To get the ideal response on the bond/gold pairing, they have to be against the same currency, which is why I'm arguing that the PP works best with domestic assets.
Interesting discussion.

I doubted that Canda and US will behave very differently in terms of inflation. Just found a historical chart as below.

It presents the inflation rates of Canada and the United States (using the CPI measure) for the period 1958 through 2008. These data are monthly and measure the inflation rate for each month as the percentage increase from the same month of the previous year. Note the remarkable similarity of the inflation rates of the two countries although they differ by as much as 2 percentage points on a few occasions for very short periods.

20091
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #292
I doubted that Canda and US will behave very differently in terms of inflation. Just found a historical chart as below.
They look pretty similar on the historical chart, but there have still been pretty big differences on smaller time scales. If you look closely at the 2003 - 2006 data, you'll find a 3 year period where the USD fell significantly against CAD. At that time, US Treasuries also weakened versus Canadian ones and there was a big performance difference in the bond markets.

So yes, the chart shows that inflation is similar, but I don't think it really captures the significant performance differences that can occur when there is divergence in our markets.

Earlier tonight I estimated the PP performance for those 3 years based on using US securities (XUS, TLT etc) vs domestic (XIU, XBB for broad bond mkt). I calculated roughly 5% CAGR using US stocks and treasuries... not horrible... but it would have been closer to 10% CAGR using domestic.

So let's call that "moderate" trouble in US dollar & treasuries, and it resulted in 5% CAGR performance loss over 3 years! That's pretty significant.

What if something like that happens again, but even more severely? I think it's a plausible scenario.
 

·
Registered
Joined
·
17 Posts
Earlier tonight I estimated the PP performance for those 3 years based on using US securities (XUS, TLT etc) vs domestic (XIU, XBB for broad bond mkt). I calculated roughly 5% CAGR using US stocks and treasuries... not horrible... but it would have been closer to 10% CAGR using domestic.

So let's call that "moderate" trouble in US dollar & treasuries, and it resulted in 5% CAGR performance loss over 3 years! That's pretty significant.

What if something like that happens again, but even more severely? I think it's a plausible scenario.
May I know which tickers you were using for the backtest? I tried to replicate the comparison but found XUS only has data available since May 2013.
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #294 (Edited)
May I know which tickers you were using for the backtest? I tried to replicate the comparison but found XUS only has data available since May 2013.
I'm using stockcharts.com with some tricks. Here's that 3 year time period.

SPY:$CADUSD shows SPY valued in CAD, same as XUS performance
... 12.99% return = 4.2% CAGR for US stocks
Then I looked at TLT. I assumed perfect hedging, pretend we get same in CAD
... 2.48% return = 0.8% CAGR for US treasuries
$GOLD:$CADUSD for gold in CAD, same as MNT but ignoring ETF fees
... 39.9% return = 11.8% CAGR for gold
XSB for cash/short term bonds, the only ETF that existed back then
... 12.12% return = 3.9% CAGR for short term bonds

Domestic alternatives
XIU = 19.8% CAGR
XBB = 4.9% CAGR (the only data I have for Cdn bonds)

Average of above, the PP using American stocks and bonds = 5.2% CAGR
If you swap to domestic you get [19.8%, 4.9%, 11.8%, 3.9%] = 10.1% CAGR

I think this is a good warning of what can happen when USD & US Treasuries are weak. It's easy to forget this, because both have been strong for about a decade now.

Some people may comment that this TSX return seems outlandishly high. Not at all. All global stocks, such as MSCI EAFE, had similar performance. America's performance was far below average, while the US dollar plummeted.
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #296
Here's a chart of my version of the permanent portfolio versus VCNS and VBAL for this year so far, as of April 16. These are daily values, total return.

The 'Permanent Portfolio' shown here is what I use. It's 15% XIU, 15% ZSP, 50% XBB, 20% MNT ... my version of it, based on various ETFs that I've been using for a long time. There is one bond fund which combines short term & long term bonds.

Some observations:
  • PP is doing the best in 2020. It's actually positive YTD!
  • As expected, the higher % stocks, the worse the crash & volatility
  • The minimum points were: PP -9%, VCNS -15%, VBAL -18%
20104
 

·
Super Moderator
Joined
·
3,024 Posts
Here's a chart of my version of the permanent portfolio versus VCNS and VBAL for this year so far, as of April 16. These are daily values, total return.

The 'Permanent Portfolio' shown here is what I use. It's 15% XIU, 15% ZSP, 50% XBB, 20% MNT ... my version of it, based on various ETFs that I've been using for a long time. There is one bond fund which combines short term & long term bonds.

Some observations:
  • PP is doing the best in 2020. It's actually positive YTD!
The run up on MNT this year is definitely giving a positive result. Question though ... since the allocations are so spread out (MNT vs others) do you plan to rebalance at all?
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #298 (Edited)
The run up on MNT this year is definitely giving a positive result. Question though ... since the allocations are so spread out (MNT vs others) do you plan to rebalance at all?
My approach is to rebalance only once a year, at year end. This is to make sure that the way I'm investing is consistent with the historical backtests, since all of those were based on annual rebalancing as well.

There are certainly other ways a person can rebalance, but I saw good historical results from the annual rebalance method, including during Great Depression years, dot com crash, etc.

I also think it's best to touch the portfolio as little as possible. We saw a good example of this in February and March; the bid/ask spreads on ETFs, especially bond ETFs, became extremely wide and corporate bonds became illiquid. Anyone who rebalanced from bonds to stocks suffered a (big) loss due to this; there was quite a large "fee" to rebalance. Was it worth it?

Optimal Rebalancing – Time Horizons Vs Tolerance Bands

As a result, a study from Vanguard, by Jaconetti, Kinniry and Zilbering, basically found no material differences in outcomes for rebalancing frequencies varying from monthly to annual, once measured on a rolling period basis. When analyzed using a 60/40 portfolio going back to 1926, the researchers found that rebalancing quarterly or monthly produced no improvement in long-term risk or returns; it simply drove up the turnover rate and the number of rebalancing events (and potential transaction costs!).
 

·
Super Moderator
Joined
·
3,024 Posts
My approach is to rebalance only once a year, at year end. This is to make sure that the way I'm investing is consistent with the historical backtests, since all of those were based on annual rebalancing as well.
Just wondering .... one of the powers to rebalancing is selling high to buy the low and combine that with a large divergence between assets (MNT vs other) I wasn't sure if you'd consider it. I mean under normal gains and dips it's not a consideration (as the studies show) but with a large (say over 30%) difference I often wonder what I'd do. And of course this could be considered a market timing mistake that only history will prove right or wrong later.
 

·
Registered
Joined
·
16,333 Posts
Discussion Starter #300 (Edited)
Just wondering .... one of the powers to rebalancing is selling high to buy the low and combine that with a large divergence between assets (MNT vs other) I wasn't sure if you'd consider it.
It's true that rebalancing works great when there is a big divergence between the assets. So the question is, how big is the current divergence in historical context?

First I constructed a "divergence from target" value for each year... the sum of the magnitude of each asset's divergence from target weight. This does a reasonable job capturing the rebalancing opportunity. The higher the value, the greater the divergences. To clean up the picture I removed bars below 6%, since that's the majority of them.

Example: with my 30/50/20 targets, in 2008 the weights ended the year at 21.7 / 52.8 / 25.5 which are divergences of 8.3%, 2.8%, 5.5%, which sum to 16.6%... biggest bar

20106


Your intuition seems correct, that 2020 (so far) is a reasonably big year for rebalancing opportunity. But I look at this and think: it's an OK rebalancing situation, but the reading isn't as high as 1997, 1999, 2002, 2008, 2013.

So I don't think it's an extreme enough situation to warrant taking action now, as opposed to year end. Sticking with the year end schedule gives me the comfort of being consistent with historical backtests. From a psychological standpoint, I also suspect that sticking with an annual routine will be easier to follow through with because "there is no choice". It's a neat idea in passive indexing... deliberately removing discretion from the picture.

Currently my only fear with the Permanent Portfolio is that I will be too scared or timid, in a year like 2002 or 2008, to execute the rebalancing. Sticking with annual rebalancing (no choice, no discretion) may be the optimal strategy from a behavioural standpoint.
 
281 - 300 of 317 Posts
Top