Bond Alternatives in the time of Zero rates

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
Nahtanoj
Posts: 32
Joined: Tue Apr 04, 2017 7:01 am

Re: Bond Alternatives in the time of Zero rates

Post by Nahtanoj »

bobcat2 wrote: Sat Aug 08, 2020 10:29 am Some things to consider when real interest rates are zero or less.

. . . .

Consider deferred life annuities. Their pricing is less affected by interest rates than immediate life annuities. Consider immediate life annuities once you are over 75.

BobK
This is interesting. Is this another way of saying that the share of a deferred life annuity’s payout that is funded by mortality credits is greater than the share of an immediate life annuity’s payout that is funded by mortality credits?

Does this imply that it is “cheaper” or “more efficient” in some sense to insure against longevity risk using deferred life annuities rather than immediate annuities?

More generally, for a retiree who can fund living expenses to life expectancy without annuitization but is concerned about longevity risk if he or his spouse lives significantly beyond life expectancy, what is a sensible strategy or decision framework to use in deciding between immediate vs. deferred annuities? (Assuming the retiree and spouse are already deferring Social Security to maximize benefits.)
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives in the time of Zero rates

Post by vineviz »

Nahtanoj wrote: Sun Aug 09, 2020 9:41 pm Does this imply that it is “cheaper” or “more efficient” in some sense to insure against longevity risk using deferred life annuities rather than immediate annuities?
For sure. A deferred income annuity is much closer to "pure" longevity insurance than a SPIA is.

Nahtanoj wrote: Sun Aug 09, 2020 9:41 pm More generally, for a retiree who can fund living expenses to life expectancy without annuitization but is concerned about longevity risk if he or his spouse lives significantly beyond life expectancy, what is a sensible strategy or decision framework to use in deciding between immediate vs. deferred annuities? (Assuming the retiree and spouse are already deferring Social Security to maximize benefits.)
I'd generally recommend setting up the deferred annuity to start paying you as late as possible (e.g. age 90), foregoing any sort of "period certain" or "refund on death" riders (unless passing on the inheritance is VERY important to you), and including the largest fixed COLA adjustment you can get (typically 4% or 5%).

Starting the payments later in life and leaving out the refund riders will mean you're really hedging the longevity risk as efficiently as possible. On the flip side, including a 4% or 5% COLA in the deferred annuity typically isn't very expensive because actuarially the insurance company knows that they aren't likely to have many years of compounding increases. And if you DO live past age 90, your chances of needing expensive long-term care or health care rise dramatically with each year.

The main risk you're taking with any sort of annuity is inflation risk. A 70-year old man might be able to buy a deferred annuity for $50k that stars paying $25k/year starting at age 90 with 4% COLA. But $25k in 2040 will buy a lot less than $25k today, so you need to adjust your intended payment to account for expected inflation.

Another solution is to purchase a single long-term TIPS bond now with the intent to purchase the deferred annuity a little later down the line. For instance, buying the same deferred annuity at age 80 instead of buying it at age 70 would reduce your inflation risk quite a bit if you "pre-fund" the purchase with a TIPS or TIPS fund that matches the time horizon of the deferred annuity.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

vineviz wrote: Sun Aug 09, 2020 4:32 pm
Northern Flicker wrote: Sun Aug 09, 2020 3:44 pm
vineviz wrote: Sun Aug 09, 2020 3:25 pm
Northern Flicker wrote: Sun Aug 09, 2020 3:04 pm
I don't get imagined returns very often. A 20% drop as the base case for real returns over say 10 years might well happen with long-term bonds, maybe it could happen with shorter duration bonds, but it is without historical precedent for short-term and intermediate term bonds.
Short-term yields are considerably more volatile than long-term yields, so broadly speaking a 20% drop in the yield on 2-year Treasuries is more likely than a similar drop in 20-year Treasuries.

Also, historically the yield curve has remained upward sloping even when real rates are negative. There are powerful theoretical reasons suggesting this should remain the case.
A 20% drop in the yield is different from a -20% real return of the asset.
Perhaps you meant to to specify a real return of -20% instead of a 20% drop in yields as the base case for estimating real returns?
No, I was addressing the position of another poster who was imagining a 20% drop in real value for bonds.
Last edited by Northern Flicker on Mon Aug 10, 2020 12:21 pm, edited 1 time in total.
Risk is not a guarantor of return.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives in the time of Zero rates

Post by vineviz »

Northern Flicker wrote: Mon Aug 10, 2020 12:02 am
No, I was addressing the position of another poster who was imaging a 20% drop in real value for bonds.
Thank you. I misinterpreted you.

Sorry for the mistake.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Nahtanoj
Posts: 32
Joined: Tue Apr 04, 2017 7:01 am

Re: Bond Alternatives in the time of Zero rates

Post by Nahtanoj »

vineviz wrote: Sun Aug 09, 2020 10:23 pm
Nahtanoj wrote: Sun Aug 09, 2020 9:41 pm Does this imply that it is “cheaper” or “more efficient” in some sense to insure against longevity risk using deferred life annuities rather than immediate annuities?
For sure. A deferred income annuity is much closer to "pure" longevity insurance than a SPIA is.

Nahtanoj wrote: Sun Aug 09, 2020 9:41 pm More generally, for a retiree who can fund living expenses to life expectancy without annuitization but is concerned about longevity risk if he or his spouse lives significantly beyond life expectancy, what is a sensible strategy or decision framework to use in deciding between immediate vs. deferred annuities? (Assuming the retiree and spouse are already deferring Social Security to maximize benefits.)
I'd generally recommend setting up the deferred annuity to start paying you as late as possible (e.g. age 90), foregoing any sort of "period certain" or "refund on death" riders (unless passing on the inheritance is VERY important to you), and including the largest fixed COLA adjustment you can get (typically 4% or 5%).

Starting the payments later in life and leaving out the refund riders will mean you're really hedging the longevity risk as efficiently as possible. On the flip side, including a 4% or 5% COLA in the deferred annuity typically isn't very expensive because actuarially the insurance company knows that they aren't likely to have many years of compounding increases. And if you DO live past age 90, your chances of needing expensive long-term care or health care rise dramatically with each year.

The main risk you're taking with any sort of annuity is inflation risk. A 70-year old man might be able to buy a deferred annuity for $50k that stars paying $25k/year starting at age 90 with 4% COLA. But $25k in 2040 will buy a lot less than $25k today, so you need to adjust your intended payment to account for expected inflation.

Another solution is to purchase a single long-term TIPS bond now with the intent to purchase the deferred annuity a little later down the line. For instance, buying the same deferred annuity at age 80 instead of buying it at age 70 would reduce your inflation risk quite a bit if you "pre-fund" the purchase with a TIPS or TIPS fund that matches the time horizon of the deferred annuity.
Thank you. This makes sense. Starting the annuity payout as late as possible should give you more longevity insurance bang for the buck. It’s also consistent with the rule of thumb that it’s often cheaper to self-insure (recognizing that some say that term is a misnomer) for costs that you can afford to pay out of pocket and buy insurance only for costs that would be difficult or impossible to pay out of pocket.

On the other hand, if you’re buying the annuity in part to protect against the risk of cognitive decline, or to make life easier for a surviving spouse who has no interest in managing even a simple portfolio of mutual funds (or even to make life easier for yourself), then you might want the payouts to start earlier. That may be behind Bob K.’s suggestion to consider immediate annuities purchased after age 75.

As for the risk of unexpected inflation, the strategy of pre-funding the annuity purchase with TIPS that are duration-matched to the estimated duration of the annuity payout (assuming you can estimate that) seems like a good one - although the real yield on long-term TIPS is currently around negative .40% or negative .50%, which is also not super-attrractive.
User avatar
patrick013
Posts: 3015
Joined: Mon Jul 13, 2015 7:49 pm

Re: Bond Alternatives in the time of Zero rates

Post by patrick013 »

Nahtanoj wrote: Mon Aug 10, 2020 8:08 am
On the other hand, if you’re buying the annuity in part to protect against the risk of cognitive decline, or to make life easier for a surviving spouse who has no interest in managing even a simple portfolio of mutual funds (or even to make life easier for yourself), then you might want the payouts to start earlier. That may be behind Bob K.’s suggestion to consider immediate annuities purchased after age 75.


If you can defer starting payments on your SPIA for ten years or longer the annual payout
can be very very good. SPIA's used to pay over $700 a month for a $100,000 investment.
Today it's hard to get $500 per month without deferral. But the tax exclusion ratio is
still about 86%. Doesn't pay quite as good as several years ago.
age in bonds, buy-and-hold, 10 year business cycle
Blackbird79
Posts: 8
Joined: Mon Aug 10, 2020 9:24 am

Re: Bond Alternatives in the time of Zero rates

Post by Blackbird79 »

I know most aren’t fans, but preferreds can be a small component if you choose wisely I’ve found the best time is during a short term liquidity crunch (earlier this year for example) when prices dip below par. Banks are loathe to cut the dividends on these vs common (look at 2008), and they’re more tax efficient.
hudson
Posts: 3237
Joined: Fri Apr 06, 2007 9:15 am

Re: Bond Alternatives in the time of Zero rates

Post by hudson »

hudson wrote: Sun Aug 09, 2020 4:58 pm For those seriously considering bond alternatives, Larry Swedroe's book, The Only Guide to Alternative Investments You'll Ever Need: The Good, the Flawed, the Bad, and the Ugly (Bloomberg Book 42), might be worth a look:

https://www.amazon.com/Only-Guide-Alter ... B003NE61GC
Here's a list of the good, flawed, bad, and ugly from Larry Swedroe's book....from a Boglehead 2007 contribution:

viewtopic.php?p=23364#p23364
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives?

Post by Northern Flicker »

Taylor Larimore wrote: Sat Aug 08, 2020 11:00 am Bogleheads:

For those searching for "Bond Alternatives," this recent Morningstar article may have the answer:

What's The Best Diversifier For Stocks?

Best wishes.
Taylor
Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
Risk is not a guarantor of return.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

Northern Flicker wrote: Mon Aug 10, 2020 12:40 pm Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
None of those are the MOST reliable diversifier of equity risk: that distinction belongs to long-term US Treasuries.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

Blackbird79 wrote: Mon Aug 10, 2020 9:40 am I know most aren’t fans, but preferreds can be a small component if you choose wisely I’ve found the best time is during a short term liquidity crunch (earlier this year for example) when prices dip below par. Banks are loathe to cut the dividends on these vs common (look at 2008), and they’re more tax efficient.
Financial preferreds were down about 64% peak to trough in 2008/2009. They will not make the list of reliable equity diversifiers:

http://quotes.morningstar.com/chart/fun ... A%5B%5D%7D
Risk is not a guarantor of return.
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives?

Post by Northern Flicker »

vineviz wrote: Mon Aug 10, 2020 12:45 pm
Northern Flicker wrote: Mon Aug 10, 2020 12:40 pm Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
None of those are the MOST reliable diversifier of equity risk: that distinction belongs to long-term US Treasuries.
That's what I migh have thought as well, but the data has not born that out.
Risk is not a guarantor of return.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

Northern Flicker wrote: Mon Aug 10, 2020 12:53 pm
vineviz wrote: Mon Aug 10, 2020 12:45 pm
Northern Flicker wrote: Mon Aug 10, 2020 12:40 pm Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
None of those are the MOST reliable diversifier of equity risk: that distinction belongs to long-term US Treasuries.
That's what I migh have thought as well, but the data has not born that out.
Of course it has. In fact, it’s not even a close call.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
000
Posts: 2723
Joined: Thu Jul 23, 2020 12:04 am

Re: Bond Alternatives?

Post by 000 »

Northern Flicker wrote: Mon Aug 10, 2020 12:40 pm
Taylor Larimore wrote: Sat Aug 08, 2020 11:00 am Bogleheads:

For those searching for "Bond Alternatives," this recent Morningstar article may have the answer:

What's The Best Diversifier For Stocks?

Best wishes.
Taylor
Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
I believe the article also mentioned that they may not do as well in the future:
Christine Benz at morningstar.com wrote: As with all backward-looking statistics, correlations data won't necessarily foretell the future: A relationship that has held in the past--high-quality bonds zigging while stocks are zagging--may not hold up. Indeed, in a recent outlook report, BlackRock Investment Institute argued that Treasuries may be less reliable as equity ballast in the future. The researchers pointed to lower starting yields as the explanation, noting that German and Japanese bonds performed less well during those countries' recent stock-market sell-offs for this very reason.
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

Stinky wrote: Sun Aug 09, 2020 3:17 pm
Northern Flicker wrote: Sun Aug 09, 2020 3:04 pm
whodidntante wrote: Sun Aug 09, 2020 11:55 am
Northern Flicker wrote: Sun Aug 09, 2020 2:57 am It was a serious question. It was not whether bonds can lose value. The phrase "bear market" conjures up the motion of an asset class bloodbath. A bear market for stocks today is defined (arbitrarily) by the financial press as a drop of 20% or more from the peak. Treasuries or the total bond index have never seen such a drawdown-- nowhere even close.

The closest thing you can find is the loss in real terms during periods of robust inflation like the 1970's. And even that was a lengthy process so that it was not so severe for short and intermediate durations.
I'm glad you mentioned real returns. I don't need a lot of imagination to consider a 20% drop on a real basis. In fact, I consider that the base case. I.e., I expect to accumulate losses on a real basis by holding high quality bonds.
I don't get imagined returns very often. A 20% drop as the base case for real returns over say 10 years might well happen with long-term bonds, maybe it could happen with shorter duration bonds, but it is without historical precedent for short-term and intermediate term bonds.

A basket of goods costing $100 in 1969 would have costed $262.94 in 1982, per this inflation calculator.

A $100 slice of 3-month rolling t-bills in 1969 with reinvestment of interest was worth $298.92 in 1982 per this historical data.
There would have been income taxes on the $198.92 of “gain” on the T-bills. So T-bills, net of taxes, wouldn’t have kept up with price inflation.
Because tax brackets are indexed to inflation. It is not clear that they would not have kept up with inflation. It would take a very detailed calculation to confirm either way.
Risk is not a guarantor of return.
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives?

Post by Northern Flicker »

vineviz wrote: Mon Aug 10, 2020 12:56 pm
Northern Flicker wrote: Mon Aug 10, 2020 12:53 pm
vineviz wrote: Mon Aug 10, 2020 12:45 pm
Northern Flicker wrote: Mon Aug 10, 2020 12:40 pm Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
None of those are the MOST reliable diversifier of equity risk: that distinction belongs to long-term US Treasuries.
That's what I migh have thought as well, but the data has not born that out.
Of course it has. In fact, it’s not even a close call.
Not in the correlation matrices in the M* article, which was surprising because it only looked back up to 15 years. But long-term treasuries were an atrocious diversifier of equity risk in the 1970's. Short and intermediate term treasuries work well across a wider range of conditions. Who knows moving forward?
Last edited by Northern Flicker on Mon Aug 10, 2020 1:15 pm, edited 1 time in total.
Risk is not a guarantor of return.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

000 wrote: Mon Aug 10, 2020 12:56 pm
Christine Benz at morningstar.com wrote: As with all backward-looking statistics, correlations data won't necessarily foretell the future: A relationship that has held in the past--high-quality bonds zigging while stocks are zagging--may not hold up. Indeed, in a recent outlook report, BlackRock Investment Institute argued that Treasuries may be less reliable as equity ballast in the future. The researchers pointed to lower starting yields as the explanation, noting that German and Japanese bonds performed less well during those countries' recent stock-market sell-offs for this very reason.
That's just a muddle of a paragraph. I don't know what the BlackRock report actually said but "starting yields" have nothing to do with how well Treasuries diversify equity risk. Now of course it's true that, all things equal, you'd prefer the return on whatever you're using for diversification to be as high as possible. It's also a bit of a truism (who wouldn't prefer higher returns over lower returns if everything else is equal?).

However diversification is purely a mathematical function of portfolio weights, correlations, and variances. Treasury bonds always have lower correlations with stocks than corporate bonds, because corporate bonds have credit risk that is directly tied to the market risk of stocks. Long-term Treasury bonds always have more variance than short- or intermediate-term Treasury bonds, because their duration is definitionally higher. Ergo, Treasury bonds have been and will be better diversifiers than shorter term Treasuries or corporate bonds of any duration unless and until the US government is less creditworthy than corporations are.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives?

Post by Northern Flicker »

000 wrote: Mon Aug 10, 2020 12:56 pm
Northern Flicker wrote: Mon Aug 10, 2020 12:40 pm
Taylor Larimore wrote: Sat Aug 08, 2020 11:00 am Bogleheads:

For those searching for "Bond Alternatives," this recent Morningstar article may have the answer:

What's The Best Diversifier For Stocks?

Best wishes.
Taylor
Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
I believe the article also mentioned that they may not do as well in the future:
Christine Benz at morningstar.com wrote: As with all backward-looking statistics, correlations data won't necessarily foretell the future: A relationship that has held in the past--high-quality bonds zigging while stocks are zagging--may not hold up. Indeed, in a recent outlook report, BlackRock Investment Institute argued that Treasuries may be less reliable as equity ballast in the future. The researchers pointed to lower starting yields as the explanation, noting that German and Japanese bonds performed less well during those countries' recent stock-market sell-offs for this very reason.
The article doesn't mention it, but this was longer term bonds that had difficulties. Even long-term treasuries had some liquidity problems in March.

Short-term treasuries are the most liquid securities in the world at present. If they do not hold up in an equity crisis, all bets are off for any other asset class. If we have a steady rise in rates, short-term treasuries will turnover and reset their rates quickly. The problem with them is a different one-- at current yields they are a drag on portfolio expected return, and their expected real return is negative.
Risk is not a guarantor of return.
000
Posts: 2723
Joined: Thu Jul 23, 2020 12:04 am

Re: Bond Alternatives?

Post by 000 »

Northern Flicker wrote: Mon Aug 10, 2020 1:12 pm
000 wrote: Mon Aug 10, 2020 12:56 pm
Northern Flicker wrote: Mon Aug 10, 2020 12:40 pm
Taylor Larimore wrote: Sat Aug 08, 2020 11:00 am Bogleheads:

For those searching for "Bond Alternatives," this recent Morningstar article may have the answer:

What's The Best Diversifier For Stocks?

Best wishes.
Taylor
Takeaways: Short and intermediate treasuries are still the most reliable diversifiers of equity risk. The total bond market index has been ok, if not fabulous for this purpose in the past, but fell down some in this role in March.
I believe the article also mentioned that they may not do as well in the future:
Christine Benz at morningstar.com wrote: As with all backward-looking statistics, correlations data won't necessarily foretell the future: A relationship that has held in the past--high-quality bonds zigging while stocks are zagging--may not hold up. Indeed, in a recent outlook report, BlackRock Investment Institute argued that Treasuries may be less reliable as equity ballast in the future. The researchers pointed to lower starting yields as the explanation, noting that German and Japanese bonds performed less well during those countries' recent stock-market sell-offs for this very reason.
The article doesn't mention it, but this was longer term bonds that had difficulties. Even long-term treasuries had some liquidity problems in March.

Short-term treasuries are the most liquid securities in the world at present. If they do not hold up in an equity crisis, all bets are off for any other asset class. If we have a steady rise in rates, short-term treasuries will turnover and reset their rates quickly. The problem with them is a different one-- at current yields they are a drag on portfolio expected return, and their expected real return is negative.
I recall the T bill market freezing up for a few days in March, don't you...?
000
Posts: 2723
Joined: Thu Jul 23, 2020 12:04 am

Re: Bond Alternatives?

Post by 000 »

vineviz wrote: Mon Aug 10, 2020 1:09 pm
000 wrote: Mon Aug 10, 2020 12:56 pm
Christine Benz at morningstar.com wrote: As with all backward-looking statistics, correlations data won't necessarily foretell the future: A relationship that has held in the past--high-quality bonds zigging while stocks are zagging--may not hold up. Indeed, in a recent outlook report, BlackRock Investment Institute argued that Treasuries may be less reliable as equity ballast in the future. The researchers pointed to lower starting yields as the explanation, noting that German and Japanese bonds performed less well during those countries' recent stock-market sell-offs for this very reason.
That's just a muddle of a paragraph. I don't know what the BlackRock report actually said but "starting yields" have nothing to do with how well Treasuries diversify equity risk. Now of course it's true that, all things equal, you'd prefer the return on whatever you're using for diversification to be as high as possible. It's also a bit of a truism (who wouldn't prefer higher returns over lower returns if everything else is equal?).
The investor demand curve at sufficiently low yields may be low enough to cause flight to other quality assets, i.e. there may be a lower bound that is approached in yields.
vineviz wrote: Mon Aug 10, 2020 1:09 pm However diversification is purely a mathematical function of portfolio weights, correlations, and variances. Treasury bonds always have lower correlations with stocks than corporate bonds, because corporate bonds have credit risk that is directly tied to the market risk of stocks. Long-term Treasury bonds always have more variance than short- or intermediate-term Treasury bonds, because their duration is definitionally higher. Ergo, Treasury bonds have been and will be better diversifiers than shorter term Treasuries or corporate bonds of any duration unless and until the US government is less creditworthy than corporations are.
I think this paragraph is a tautology.
Last edited by 000 on Mon Aug 10, 2020 1:23 pm, edited 1 time in total.
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

I believe it was just one day for bills, but t-bills and federally insured deposit accounts were still the safest and most liquid thing you could have held on that day.
Risk is not a guarantor of return.
000
Posts: 2723
Joined: Thu Jul 23, 2020 12:04 am

Re: Bond Alternatives in the time of Zero rates

Post by 000 »

Northern Flicker wrote: Mon Aug 10, 2020 1:19 pm I believe it was just one day for bills, but t-bills and federally insured deposit accounts were still the safest and most liquid thing you could hold on that day.
I think bank accounts beat T bills that day.
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

For a buy and hold investor, i-bonds are probably the safest fixed income investment available to US investors.
Risk is not a guarantor of return.
000
Posts: 2723
Joined: Thu Jul 23, 2020 12:04 am

Re: Bond Alternatives in the time of Zero rates

Post by 000 »

Northern Flicker wrote: Mon Aug 10, 2020 1:21 pm For a buy and hold investor, i-bonds are probably the safest fixed income investment available to US investors.
Other than the risk of not having enough when you need it :twisted:
User avatar
HomerJ
Posts: 15164
Joined: Fri Jun 06, 2008 12:50 pm

Re: Bond Alternatives in the time of Zero rates

Post by HomerJ »

Explorer wrote: Sun Jul 05, 2020 10:08 am
DanFrancis wrote: Sat Jul 04, 2020 11:26 pm You seem so certain that you would lose money in bonds over the next 6 years...I'm not sure you are correct. It's a guess...but nothing more than a guess. I'm trying to think of any 6 year period where bonds went down every year. I can't off the top of my head. Some CD's would be okay...but that is basically a form of a bond. I hate annuities...too expense for the product. I would have a balanced portfolio, not know what the future really holds.
Bold highlight is mine. As other posters said, right after world war II, bond yields started going up rapidly causing bondholders lose a heck of a lot of value on their bonds. Do not assume bonds are safe.. in the current environment they are a safe way to lose money.
This is incorrect.

Rising interest rates is not a bond fund killer. Inflation is the danger.

People talk about losing money in bonds from 1945-1982, but it was really the last 5 years of that period that caused the losses.

Double-digit inflation and hugely rising interest rates at the same time.

Yes, bonds going forward are not likely to do well, but if inflation stays fairly normal, it won't be that bad.

When interest rates go up, yes bond funds drop in value, but new bonds are purchased with the higher interest rates, and the bond fund starts paying out more each year.

It all evens out. Bond funds are pretty steady regardless of what interest rates are doing.

Inflation is the danger, and inflation is what hurt bonds in the late 1970s.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

Northern Flicker wrote: Mon Aug 10, 2020 1:01 pm Not in the correlation matrices in the M* article, which was surprising because it only looked back up to 15 years.
That'd only be relevant if correlation were the sole factor that determined diversification benefit. It's not.

All Treasuries roughly the same correlation with stocks. The benefit of long-term Treasuries as diversifiers isn't due to correlations being lower, but rather to the variance being higher. You can see here the relative diversification of various Vanguard bond funds in 60/50 combinations with the S&P 500 (VFINX = 1).

Image

Take a look at the underlying data: the Treasury funds all have correlation of roughly -0.30 for the period sampled, in which case the higher volatility asset is the better diversifier. https://www.portfoliovisualizer.com/ass ... &months=36
Last edited by vineviz on Mon Aug 10, 2020 2:11 pm, edited 1 time in total.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

Yes, I agree that during the long secular bull market for treasuries with rates falling when equities fell, LTT have been a better diversifier. But shorter duration treasuries have vern effective across a wider range of conditions. Purely for equity diversification across all conditions, intermediate treasuries have been the most effective historically. Unfortunately, we won't see this with PV given the historical bias in its data.

From a liability matching perspective, equity diversification liabilities have a much shorter expected duration than the liabilities associated with retiring at a typical retirement age.
Risk is not a guarantor of return.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

000 wrote: Mon Aug 10, 2020 1:17 pm
vineviz wrote: Mon Aug 10, 2020 1:09 pm However diversification is purely a mathematical function of portfolio weights, correlations, and variances. Treasury bonds always have lower correlations with stocks than corporate bonds, because corporate bonds have credit risk that is directly tied to the market risk of stocks. Long-term Treasury bonds always have more variance than short- or intermediate-term Treasury bonds, because their duration is definitionally higher. Ergo, Treasury bonds have been and will be better diversifiers than shorter term Treasuries or corporate bonds of any duration unless and until the US government is less creditworthy than corporations are.
I think this paragraph is a tautology.
As I said, the very definition of diversification is a strict mathematical function with only three inputs. If two bond funds have the same correlation to stocks, it's impossible for the lower volatility fund to be the better diversifier.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives in the time of Zero rates

Post by vineviz »

Northern Flicker wrote: Mon Aug 10, 2020 2:00 pm Yes, I agree that during the long secular bull market for treasuries with rates falling when equities fell, LTT have been a better diversifier.
The bull market has nothing to do with anything: the return of the bond fund is not an input into the calculation of diversification benefit.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

000 wrote: Mon Aug 10, 2020 1:22 pm
Northern Flicker wrote: Mon Aug 10, 2020 1:21 pm For a buy and hold investor, i-bonds are probably the safest fixed income investment available to US investors.
Other than the risk of not having enough when you need it :twisted:
That's why you hold equities. Hold stocks so you can eat well. Hold bonds so you can sleep well.
Risk is not a guarantor of return.
reln
Posts: 446
Joined: Fri Apr 19, 2019 4:01 pm

Re: Brighthouse Shield Annuities.

Post by reln »

ruralavalon wrote: Sun Aug 09, 2020 4:47 pm
reln wrote: Sun Aug 09, 2020 3:28 pm
Compass wrote: Sat Jul 04, 2020 6:19 pm I'm considering using these products in my traditional IRA to replace a portion, not all, of my bond mutual funds. These got a bad rap on our forum when comparing them to no load index stock mutual funds mostly because they don't pay dividends. But the commenters did not discuss them as an alternative to bond mutual funds. With interest rates near zero, any bond mutual funds probably will loose within 6 years as rates can only go up. This seems like it could solve that problem by transferring all of the bonds in a my traditional IRA to a combination of CDs and this annuity. With Bond Mutual funds posed to loose over the next 6 years, the strategy is to reallocate the to these locked up term Shield annuities for some stability and some returns and keep some CDs/cash to re-balance with when stocks crash.

There are 2 Shield products in particular. 1st is the 6 year term Sheild 25 with a 125% Cap based on the S&P index this is recommended for stability. The 2nd is the 1 year term Shield 10 with over a 9% Step return. This one is more of a gamble but with decent odds.

I'd appreciate your opinion on this strategy on using these Brighthouse Shield Annuity contracts, not as an alternative to Stock mutual funds, but as as an alternative to Bonds with a mix of cash.
I wouldn't do it myself but index annuities are reasonable alternatives to bond funds.

I also wouldn't hold bonds funds.
Indexed annuities (aka equity indexed annuities) are not a reasonable alternative to a bond fund. They are not a reasonable investment vehicle at all in my opinion. They are complex and there are cons to consider, such as high fees and commissions that are often associated with them.

FINRA, "Indexed annuities" .

FINRA, An alert on equity-indexed annuities.
I disagree with your conclusion.

But I hold neither index annuities or bond funds, so I don't care what your conclusion is.
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Bond Alternatives?

Post by Steve Reading »

000 wrote: Mon Aug 10, 2020 1:17 pm I think this paragraph is a tautology.
The problem is that vineviz is using a specific definition of "diversification". One that only popped up in 2008, created by Choueifaty and Coignard and I think even trademarked. Because vineviz is so adamant this is how diversification is defined, then perhaps he believes this definition superseded the previous definition (since "diversification" certainly existed in financial vocabulary pre-2008).

It's a peculiar definition because it's very different than what writers like Bernstein and Swedroe mean when they say "diversification". Diversification (as per these other writers) is called "a free lunch" because you can increase returns and lower risk at the same time, leading to higher risk-adjusted returns. It is this diversification that gave Ray Dalio the light bulb moment to start Pure Alpha decades ago (if you're familiar with that story). When you hear "diversification is a free lunch", this is the definition they mean.

The 2008 definition vineviz uses is a specific mathematical definition that does not necessarily lead to higher risk-adjusted returns. By the metric, adding inverse ETFs offer amazing diversification! Very volatile and almost -1.0 correlation. But it's not actually improving risk-adjusted returns and hence, is no free lunch.

The confusion comes when different posters use different definitions. Here, Northern Flicker claims IT treasuries are a better "diversifier" and I know what he means is that "they're a better complement to equities for portfolios". He believes he will get better risk-adjusted returns via these treasuries. Here, diversification is always good (provided the risk-return is appropriate to you).
Northern Flicker wrote: Mon Aug 10, 2020 1:01 pm Not in the correlation matrices in the M* article, which was surprising because it only looked back up to 15 years. But long-term treasuries were an atrocious diversifier of equity risk in the 1970's. Short and intermediate term treasuries work well across a wider range of conditions. Who knows moving forward?
But when vineviz says it, you actually can't conclude anything about whether it's worth using IT or LT treasuries. Yes, it is "more diversified" (by 2008 definition) but your more diversified portfolio could end up with lower returns and higher risks as a result.

Hope that gives a bit more context.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
patrick
Posts: 1791
Joined: Fri Sep 04, 2009 3:39 am
Location: Mega-City One

Re: Brighthouse Shield Annuities.

Post by patrick »

Compass wrote: Sat Jul 04, 2020 6:19 pm I'm considering using these products in my traditional IRA to replace a portion, not all, of my bond mutual funds. These got a bad rap on our forum when comparing them to no load index stock mutual funds mostly because they don't pay dividends.
Throwing away dividends isn't the only reason against equity indexed annuities. They also typically cap your gains, charge withdrawal penalties if you don't hold them for many years, and allow the insurance company to adjust the cap every year.
But the commenters did not discuss them as an alternative to bond mutual funds. With interest rates near zero, any bond mutual funds probably will loose within 6 years as rates can only go up. This seems like it could solve that problem by transferring all of the bonds in a my traditional IRA to a combination of CDs and this annuity. With Bond Mutual funds posed to loose over the next 6 years, the strategy is to reallocate the to these locked up term Shield annuities for some stability and some returns and keep some CDs/cash to re-balance with when stocks crash.

There are 2 Shield products in particular. 1st is the 6 year term Sheild 25 with a 125% Cap based on the S&P index this is recommended for stability. The 2nd is the 1 year term Shield 10 with over a 9% Step return. This one is more of a gamble but with decent odds.

I'd appreciate your opinion on this strategy on using these Brighthouse Shield Annuity contracts, not as an alternative to Stock mutual funds, but as as an alternative to Bonds with a mix of cash.
This product seems (from my brief look) rather different from the typical equity indexed annuity. It only absorbs a limited amount of market decline so you still could lose a lot if the market goes down too much, making this product a poor bond alternative.

Beyond that, bond funds might not lose over the next 6 years. If interest rates stay where they are, bond funds will gain a little bit since rates are slightly above zero. You may think interest rates must rise soon, but people thought that 6 years ago too, and were wrong. Perhaps whatever is keeping interest rates low will continue.

Bonds losing money after inflation is more likely, but this annuity seems to lack a guarantee of keeping up with inflation. If interest rate do surge, this annuity still would not be ideal. You would have to wait until the surrender period is up (or else pay the penalty) before you could move the money over to higher-yielding bonds.
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

vineviz wrote: Mon Aug 10, 2020 2:11 pm
Northern Flicker wrote: Mon Aug 10, 2020 2:00 pm Yes, I agree that during the long secular bull market for treasuries with rates falling when equities fell, LTT have been a better diversifier.
The bull market has nothing to do with anything: the return of the bond fund is not an input into the calculation of diversification benefit.
No, but the fact that LTT got hammered with stocks in the 1970's is relevant. The point of my comment was not the return component but that it was a long period of disinflation and mild deflation for very short periods during ewuity downturns. These are conditions in which LTT are a better diversifier of equities than short or intermediate treasuries.

During accelerating inflation, LTT have as good a job of this. The problem is exacerbated by the market for LTT''s being driven by large institutional investirs like pensions and insurance companies that are matching them to nominal liabilities. Because they are as a result not taking inflation risk, LTT's generally have yields that underprice inflation risk over their term, undermining their ability to serve as a good equity diversifier during periods of robust inflation. LTT's do not match long-term real liabilities very well, creating substantial risk for retirement savers who maintain long duration nominal bond portfolios as their fixed income portfolio.
Risk is not a guarantor of return.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives in the time of Zero rates

Post by vineviz »

Northern Flicker wrote: Mon Aug 10, 2020 3:31 pm LTT's do not match long-term real liabilities very well, creating substantial risk for retirement savers who maintain long duration nominal bond portfolios as their fixed income portfolio.
Clearly you can see how backwards this sentence is?

Long-term (nominal) Treasuries match long-term real liabilities better than any other asset apart from long-term TIPS. There's no way to argue that a short-term nominal bond is a better match for long-term liabilities than a long-term bond is.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

Steve Reading wrote: Mon Aug 10, 2020 2:49 pm It's a peculiar definition because it's very different than what writers like Bernstein and Swedroe mean when they say "diversification". Diversification (as per these other writers) is called "a free lunch" because you can increase returns and lower risk at the same time, leading to higher risk-adjusted returns. It is this diversification that gave Ray Dalio the light bulb moment to start Pure Alpha decades ago (if you're familiar with that story). When you hear "diversification is a free lunch", this is the definition they mean.
It only seems "peculiar" because so many of these writers use the word "diversification" to refer to something other than diversification, typically by introducing return into the discussion. At that point, often, they're drawing a false synonymity between a "diversified" portfolio and a "mean-variance optimal" portfolio.
Steve Reading wrote: Mon Aug 10, 2020 2:49 pmBy the metric, adding inverse ETFs offer amazing diversification! Very volatile and almost -1.0 correlation. But it's not actually improving risk-adjusted returns and hence, is no free lunch.

As you know, trying to diversify an asset with itself is completely nonsensical. It's the equivalent of dividing 0 by 0 in arithmetic: every answer is wrong because the very expression is undefined.

In the case of nominal Treasuries the answer is both obvious and intuitive: All three duration classes (long, intermediate, and short) have statistically indistinguishable correlations with equities, pure exposure to the same risk factor (i.e term risk) in various amounts, and volatilities that are in direct proportion with the risk exposures. It's the simplest possible diversification problem to solve, and there's no way to get to any answer OTHER than LTTs being the better diversifier without making contrafactual assumptions.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
caklim00
Posts: 2271
Joined: Mon May 26, 2008 10:09 am

Re: Bond Alternatives in the time of Zero rates

Post by caklim00 »

I've been wondering whether my strategy of holding Treasury Futures long term as effectively adding bond exposure to my mostly equity exposure is going to end up being a bad strategy. So far my Treasury futures ladder is up around 50K over the past 6 months. Not sure how much longer this can work. At some point PSLDX, Hedgefundie, etc. has to blow up right?
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Bond Alternatives?

Post by Steve Reading »

vineviz wrote: Mon Aug 10, 2020 4:05 pm
Steve Reading wrote: Mon Aug 10, 2020 2:49 pm It's a peculiar definition because it's very different than what writers like Bernstein and Swedroe mean when they say "diversification". Diversification (as per these other writers) is called "a free lunch" because you can increase returns and lower risk at the same time, leading to higher risk-adjusted returns. It is this diversification that gave Ray Dalio the light bulb moment to start Pure Alpha decades ago (if you're familiar with that story). When you hear "diversification is a free lunch", this is the definition they mean.
It only seems "peculiar" because so many of these writers use the word "diversification" to refer to something other than diversification, typically by introducing return into the discussion. At that point, often, they're drawing a false synonymity between a "diversified" portfolio and a "mean-variance optimal" portfolio.
I find it peculiar because the definition that was used for decades before had very specific, actionable properties (i.e. "it's generally best to diversify") that don't apply any more with this alternative definition. If a portfolio is more diversified using the 2008 definition, it might have higher risks and lower returns. So diversification isn't always a good thing using that definition. I find that peculiar.
vineviz wrote: Mon Aug 10, 2020 4:05 pm As you know, trying to diversify an asset with itself is completely nonsensical. It's the equivalent of dividing 0 by 0 in arithmetic: every answer is wrong because the very expression is undefined.
Totally nonsensical. Which is why I shy away from a measure of diversification would score that so highly. I think you missed my broader point. It's not robust to assets having negative factor loadings to one another. This isn't specific to shorts obviously. You can't measure "diversification" for value and momentum funds using that definition, for instance, because they each contain a small "short" on each other.
vineviz wrote: Mon Aug 10, 2020 4:05 pm In the case of nominal Treasuries the answer is both obvious and intuitive: All three duration classes (long, intermediate, and short) have statistically indistinguishable correlations with equities, pure exposure to the same risk factor (i.e term risk) in various amounts, and volatilities that are in direct proportion with the risk exposures. It's the simplest possible diversification problem to solve, and there's no way to get to any answer OTHER than LTTs being the better diversifier without making contrafactual assumptions.
In this scenario, the 2008 definition will not make you short things, because you've specifically eliminated the assets that would be silly to allocate to (but which it would allocate to if left by itself). But whatever the result (which would be LTTs+equities) might have lower returns AND higher risk. By the 2008 definition, it is more diversified.

One just has to be careful, that's all. Using the 2008 definition, it is not a free lunch to diversify necessarily. It actually might be a free lunch to un-diversify. So whatever results you get might have not bearing on what you actually should do.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

vineviz wrote: Mon Aug 10, 2020 3:46 pm
Northern Flicker wrote: Mon Aug 10, 2020 3:31 pm LTT's do not match long-term real liabilities very well, creating substantial risk for retirement savers who maintain long duration nominal bond portfolios as their fixed income portfolio.
Clearly you can see how backwards this sentence is?

Long-term (nominal) Treasuries match long-term real liabilities better than any other asset apart from long-term TIPS. There's no way to argue that a short-term nominal bond is a better match for long-term liabilities than a long-term bond is.
I never said that short-term treasuries were a better match than LTT to long duration real liabilities. I said short and intermediate treasuries were better equity diversifiers. I also said that LTT were not a good match to long duration real liabilities. If your goal is to match to long-term real liabilities, long-term TIPS are the bonds that do that.

As far as whether short-term treasuries or long-term treasuries better match long duration expected real liabilities, it is not as easy a question to answer as you suggest. It requires, among other things, knowing the probability distribution of future inflation.

Inflation risk and the mismatch of long-term nominal bonds with long duration real liabilities are things I believe you have too readily dismissed when you make regular recommendations of long-term nominal treasuries to retirement savers.
Last edited by Northern Flicker on Mon Aug 10, 2020 7:27 pm, edited 1 time in total.
Risk is not a guarantor of return.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

Steve Reading wrote: Mon Aug 10, 2020 4:48 pm I find it peculiar because the definition that was used for decades before had very specific, actionable properties (i.e. "it's generally best to diversify") that don't apply any more with this alternative definition. If a portfolio is more diversified using the 2008 definition, it might have higher risks and lower returns. So diversification isn't always a good thing using that definition. I find that peculiar.
The definition isn't actually new to 2008 (though the express derivation of the so-called diversifcation ratio dates to that year): it's called Markowitz diversification for a reason, after all.

Furthermore, while the force of diversification is - itself - agnostic to returns would take a set of very odd assets (or at least an odd set of relative risk/return assumptions about those assets) to create the result you're hypothesizing (improving diversification, lowering return, AND increasing volatility) and it would take a pretty weak portfolio construction process to accept that result as a "good thing".

Traditional mean-variance optimization (for all it's drawbacks) already gives us a framework for deciding between portfolios the high risk/low return and low risk/high return: we don't need diversification to inform that decision. Instead, quantifying diversification provides a third, orthogonal dimension to a framework that has long suffered from being viewed simply as a two-dimensional problem.

I'm not sure what the objection is to having three characteristics of a portfolio (expected return, expected volatility, and diversification) instead of just two but it seems very reasonable and actually desirable. One problem with tradition al MVO is that it will - if left unconstrained - generate highly concentrated (i.e. under diversified) portfolios which are highly sensitive to the return assumptions. Markowtiz diversification represents, IMHO, a reasonable constraint to apply to MVO. Or vice versa. Especially if we are operating under leverage restrictions, since the MVO portfolio is often quite low in both the risk and return dimensions.

Steve Reading wrote: Mon Aug 10, 2020 4:48 pm One just has to be careful, that's all. Using the 2008 definition, it is not a free lunch to diversify necessarily. It actually might be a free lunch to un-diversify. So whatever results you get might have not bearing on what you actually should do.
The free lunch is still there: it's just quantified, instead of subjectively evaluated. A diversified portfolio will, under Markowitz, combine assets that produce a portfolio with a return that is greater the weighted average return of the assets AND a volatility that is less than the weighted average volatility of the assets. That's the free lunch of diversification, at least under CAPM, and it's also the police force so to speak: valuing diversification as a distinct vector from return and volatility provides the portfolio architect with a way to optimize MV without loading up the portfolio with highly concentrated bets on idiosyncratic risks.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Bond Alternatives?

Post by Steve Reading »

vineviz wrote: Mon Aug 10, 2020 6:13 pm The definition isn't actually new to 2008 (though the express derivation of the so-called diversifcation ratio dates to that year): it's called Markowitz diversification for a reason, after all.
Oh I didn't know that mathematical definition of diversification predated Choueifaty and Coignard and 2008. I have never seen it in any writing except that paper. Do you have a source for this? I genuinely would like to read more about it in earlier contexts.
vineviz wrote: Mon Aug 10, 2020 6:13 pm it's called Markowitz diversification for a reason, after all.
Are you implying Markowitz came up with it? Didn't even know it was called that. Was it just simply a homage by Choueifaty and Coignard when they introduced the concept in 2008.

You know, how Ayres and Nalebuff call their rule the "Samuelson Share" in their lifecycle investing book.
vineviz wrote: Mon Aug 10, 2020 6:13 pm Furthermore, while the force of diversification is - itself - agnostic to returns would take a set of very odd assets (or at least an odd set of relative risk/return assumptions about those assets) to create the result you're hypothesizing (improving diversification, lowering return, AND increasing volatility)
Commodities, factor funds and even LT treasuries are all capable of producing more "diversified" portfolios by the 2008 definition, while actually having lower returns and even higher risks. In fact, any time you have assets that don't have very similar Sharpes, diversification will output portfolios that are inefficient (some times grossly inefficient). You might consider these "very odd assets" but I don't.
vineviz wrote: Mon Aug 10, 2020 6:13 pm Traditional mean-variance optimization (for all it's drawbacks) already gives us a framework for deciding between portfolios the high risk/low return and low risk/high return: we don't need diversification to inform that decision. Instead, quantifying diversification provides a third, orthogonal dimension to a framework that has long suffered from being viewed simply as a two-dimensional problem.
Diversification isn't a third, orthogonal metric. It's just MVO, assuming equal Sharpes. If you find the MVO, equal Sharpe (aka Maximum Diversification) portfolio useful as a framework, that's fine. I personally think Sharpe is a crap measure and don't use MVO much in my portfolios, if at all.
vineviz wrote: Mon Aug 10, 2020 6:13 pm I'm not sure what the objection is to having three characteristics of a portfolio (expected return, expected volatility, and diversification) instead of just two but it seems very reasonable and actually desirable. One problem with tradition al MVO is that it will - if left unconstrained - generate highly concentrated (i.e. under diversified) portfolios which are highly sensitive to the return assumptions. Markowtiz diversification represents, IMHO, a reasonable constraint to apply to MVO. Or vice versa. Especially if we are operating under leverage restrictions, since the MVO portfolio is often quite low in both the risk and return dimensions.
Maximum diversification IS unconstrained MVO, but with the assumption that assets have equal Sharpe. It's not a constraint on MVO. It's just MVO with some arbitrary returns on assets.
vineviz wrote: Mon Aug 10, 2020 6:13 pm The free lunch is still there: it's just quantified, instead of subjectively evaluated. A diversified portfolio will, under Markowitz, combine assets that produce a portfolio with a return that is greater the weighted average return of the assets AND a volatility that is less than the weighted average volatility of the assets.
No portfolio will have higher arithmetic return than the weighted average returns of its assets. And every portfolio will always have lower volatility than the weighted average of their assets (except in trivial cases).

The most diversified portfolio (2008 definition) maximizes on that second one. But that gives you no assurance that that's the more useful portfolio. In other words, it is true LTTs+equities is "more diversified" than ITTs+equities, but it is not necessarily true that the former will have less volatility than the latter. And it also might or might not have higher returns, depends on your return estimates.

You get this situation where your metric says "this portfolio is more diversified if you do this" but the risk of the portfolio actually goes up when you do that. And the return might or might not go up. I don't think one would consider that more diversified portfolio a "free lunch".
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
UberGrub
Posts: 139
Joined: Wed Aug 21, 2019 3:47 pm

Re: Bond Alternatives?

Post by UberGrub »

Steve Reading wrote: Mon Aug 10, 2020 7:02 pm In other words, it is true LTTs+equities is "more diversified" than ITTs+equities, but it is not necessarily true that the former will have less volatility than the latter.
Wait, is this true? I thought if you have portfolio equity+LTTs (say 60/40) and equity+ITTs (60/40 also), that because treasuries have similar ~0 correlation to stocks but LTTs are more volatile, that the first portfolio would actually have lower variance because there is a more powerful diversification effect. Am I wrong?
Northern Flicker
Posts: 6497
Joined: Fri Apr 10, 2015 12:29 am

Re: Bond Alternatives in the time of Zero rates

Post by Northern Flicker »

Currently, published correlations of US stocks and treasuries are negative. These are historical sample correlations. In the late 1980's when I started learning about investing, published sample correlations between US stocks and treasuries (the total treasury market) were about 0.6.
Last edited by Northern Flicker on Tue Aug 11, 2020 12:46 pm, edited 1 time in total.
Risk is not a guarantor of return.
User avatar
HomerJ
Posts: 15164
Joined: Fri Jun 06, 2008 12:50 pm

Re: Bond Alternatives in the time of Zero rates

Post by HomerJ »

Never ask an economist for investing advice.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Bond Alternatives?

Post by Steve Reading »

UberGrub wrote: Mon Aug 10, 2020 7:20 pm
Steve Reading wrote: Mon Aug 10, 2020 7:02 pm In other words, it is true LTTs+equities is "more diversified" than ITTs+equities, but it is not necessarily true that the former will have less volatility than the latter.
Wait, is this true? I thought if you have portfolio equity+LTTs (say 60/40) and equity+ITTs (60/40 also), that because treasuries have similar ~0 correlation to stocks but LTTs are more volatile, that the first portfolio would actually have lower variance because there is a more powerful diversification effect. Am I wrong?
In the above case, the LTTs+equity will have higher variance. It probably will have higher returns too (since LTTs have higher rates than ITTs). Whether the additional return compensates for the additional portfolio variance to produce higher risk-adjusted returns depends on the specific volatility and return estimates you have of the assets.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
000
Posts: 2723
Joined: Thu Jul 23, 2020 12:04 am

Re: Bond Alternatives?

Post by 000 »

Steve Reading wrote: Mon Aug 10, 2020 2:49 pm
000 wrote: Mon Aug 10, 2020 1:17 pm I think this paragraph is a tautology.
The problem is that vineviz is using a specific definition of "diversification". One that only popped up in 2008, created by Choueifaty and Coignard and I think even trademarked. Because vineviz is so adamant this is how diversification is defined, then perhaps he believes this definition superseded the previous definition (since "diversification" certainly existed in financial vocabulary pre-2008).
Thanks. It's hard to communicate with people using their own private language.
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

000 wrote: Tue Aug 11, 2020 1:12 pm
Steve Reading wrote: Mon Aug 10, 2020 2:49 pm
000 wrote: Mon Aug 10, 2020 1:17 pm I think this paragraph is a tautology.
The problem is that vineviz is using a specific definition of "diversification". One that only popped up in 2008, created by Choueifaty and Coignard and I think even trademarked. Because vineviz is so adamant this is how diversification is defined, then perhaps he believes this definition superseded the previous definition (since "diversification" certainly existed in financial vocabulary pre-2008).
Thanks. It's hard to communicate with people using their own private language.
As I told Steve Reading earlier, the meaning of diversification didn't suddenly change in 2008. Choueifaty and Coignard didn't create the definition (just like Fama and French didn't "create" the concept of small stocks or value stocks), which is foundational in MPT.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 7815
Joined: Tue May 15, 2018 1:55 pm

Re: Bond Alternatives?

Post by vineviz »

Steve Reading wrote: Mon Aug 10, 2020 7:02 pm Diversification isn't a third, orthogonal metric. It's just MVO, assuming equal Sharpes. If you find the MVO, equal Sharpe (aka Maximum Diversification) portfolio useful as a framework, that's fine. I personally think Sharpe is a crap measure and don't use MVO much in my portfolios, if at all.
I think you focusing so much on your objections to the maximum diversification portfolio that you're not visualizing the diversification metric itself as a distinct concept. Maybe this surface plot will help me explain it more clearly.

Basically a three-dimensional mean-variance frontier, with expected volatility and expected return on the x- axis and y- axis respectively and with diversification ratio on the z-axis. The assets I used were arbitrary (and the bins were crude) but I think the concept should be clear.

Image
Steve Reading wrote: Mon Aug 10, 2020 7:02 pm Maximum diversification IS unconstrained MVO, but with the assumption that assets have equal Sharpe. It's not a constraint on MVO. It's just MVO with some arbitrary returns on assets.
Again, leave aside the MD portfolio for minute. What I'm suggesting is that the diversification ratio CAN be used as a constraint on an MVO optimizer (e.g. given set of return, volatility, and correlation assumptions generate the portfolio with the highest expected return for a given amount of volatility subject to the constraint that the diversification ratio be greater than X). Taking this approach allows you to generate a plausible portfolio with suitable risk-return characteristics while taking steps to avoid concentrating weight in the assets whose expected returns you are most overly optimistic.
Steve Reading wrote: Mon Aug 10, 2020 7:02 pm No portfolio will have higher arithmetic return than the weighted average returns of its assets. And every portfolio will always have lower volatility than the weighted average of their assets (except in trivial cases).
Right, but it's not the arithmetic mean to which I was referring. One of the most important impacts of diversification is that it helps tug the geometric mean and - most importantly - the harmonic mean back towards the arithmetic mean. This is particularly crucial for retirement portfolios (i.e. portfolios under decumulation), where the first two moments alone are not sufficient to solve for the optimal portfolio. Using DR is a bit of kludge, but it is surprisingly effective on both theoretical grounds and empirical grounds at using ex-post return data to very cleanly approximate the ex-ante SWR-optimal portfolio.

I know you don't like the implicit assumption of equal Sharpe ratios. I'm less convinced than you are that this is a problematic assumption, but I it is incredibly easy to implement a process where you are jointly optimizing the three dimensions I illustrated above (expected return, expected volatility, expected diversification) while including either implicit or explicit assumptions about any of the inputs (returns, volatility, correlations).

The key, though, is to be able to articulate that a strictly MVO portfolio is NOT necessarily the same (and often is not the same) as the maximum diversification portfolio. I know you get that distinction, and maybe it seems so obvious as to merit no discussion, but my experience is that many investors (including many professionals) are unable to articulate this. They often say, and therefore I presume they think, that the portfolios on the MV frontier are maximally diverse even though they are clearly not.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Bond Alternatives?

Post by Steve Reading »

vineviz wrote: Tue Aug 11, 2020 3:33 pm One of the most important impacts of diversification is that it helps tug the geometric mean and - most importantly - the harmonic mean back towards the arithmetic mean.
A portfolio that scores higher in the "diversification ratio" can have geometric mean that's farther away from its arithmetic mean (both in absolute and relative terms), than a portfolio that scores lower. Actually, the Minimum Variance portfolio probably stands a better chance of doing the above.

So while people like you and I would argue "hey, the more diversified the portfolio, the more the geometric mean is tugged to the arithmetic mean", the equation you use does not necessarily lead to that.
vineviz wrote: Tue Aug 11, 2020 3:33 pm Using DR is a bit of kludge, but it is surprisingly effective on both theoretical grounds and empirical grounds at using ex-post return data to very cleanly approximate the ex-ante SWR-optimal portfolio.
]The above should only be the case when the possible assets have roughly similar Sharpe. So when it was applied to individual stocks, like the 2008 paper does, it should do generally OK. If applied with very dissimilar assets (like I said before, commodities, LT treasuries and factor funds), I would not expect that it would be "surprisingly effective". I probably wouldn't expect it to be more effective than, say, the Minimum Variance portfolio.

I'd like to see some of the empirical grounds you're referring to. Do you have a paper I could look at?
vineviz wrote: Tue Aug 11, 2020 3:33 pm The key, though, is to be able to articulate that a strictly MVO portfolio is NOT necessarily the same (and often is not the same) as the maximum diversification portfolio. I know you get that distinction, and maybe it seems so obvious as to merit no discussion, but my experience is that many investors (including many professionals) are unable to articulate this. They often say, and therefore I presume they think, that the portfolios on the MV frontier are maximally diverse even though they are clearly not.
This depends on the definition used. Many would say diversification was almost akin to efficiency. The "free lunch". I've read papers by Samuelson where these terms are somewhat interchangeable in his lexicon.

I suppose the part that annoys me is that the two fellas came out in 2008, named their thing something very widespread, and now they can claim things like "this is more diversified" when it isn't more diversified by the original standard. I'm still open to any source that shows this isn't the case, and that this measure was in fact commonly used before.

(Just a joke) You and I should come up with some arbitrary MVO solution (say, MVO with correlations inversely proportional to volatility?), then call it the "most efficiently diversified", trademark it and then pop into threads saying "well, clearly you need a lot of EM, commodities and bitcoin to have the most efficiently-diversified portfolio. Obviously I'm being tongue-in-cheek. :happy
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Bond Alternatives?

Post by Steve Reading »

000 wrote: Tue Aug 11, 2020 1:12 pm
Steve Reading wrote: Mon Aug 10, 2020 2:49 pm
000 wrote: Mon Aug 10, 2020 1:17 pm I think this paragraph is a tautology.
The problem is that vineviz is using a specific definition of "diversification". One that only popped up in 2008, created by Choueifaty and Coignard and I think even trademarked. Because vineviz is so adamant this is how diversification is defined, then perhaps he believes this definition superseded the previous definition (since "diversification" certainly existed in financial vocabulary pre-2008).
Thanks. It's hard to communicate with people using their own private language.
Yes indeed!
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
Post Reply