deleted

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
ogd
Posts: 4875
Joined: Thu Jun 14, 2012 11:43 pm

Re: How low are interest rates, really?

Post by ogd » Tue Apr 29, 2014 11:30 am

Atgard wrote:Like I said, I think the risk that stocks have negative real return over a 30-year period (which is theoretically possible, but has never happened) is less than the risk that 30-year bonds with historically low rates will have negative real return over that period.

I know either, or both, or neither could happen. And, if the stock market "goes to zero," as in the article you linked, you think those treasuries will be worth anything (they may be "paid" in thoroughly-deflated dollars)? If everything collapses, everything collapses. (That's another thread, which involves guns and gold.) Like I said, I plan for the reasonably probable. If I tried to plan for every possible "six sigma" event and let that outweigh the probable outcomes, I wouldn't invest in stocks or bonds and would miss out on everything while worrying about very unlikely scenarios.
Atgard: leaving aside for a moment that I'm not convinced that history is such a strong guarantee, as opposed to the dates and amounts printed on a bond -- like I said, I don't want to be one of the victims of a historical first -- the other thing is, it's not just about the 30 year goalpoast. In the interim, that long bond is likely to be there for me if I need it and my stocks are in the dumpster. Replacing it with a stock just won't do the same, period. If 10 years from now if we're in a deep crash and I need money to live the notion that I only need to wait another 20 years, and historically ... etc, won't do anything for me or pay my bills. When the stock market went not to zero but close, 11% of its former value, the Treasury was king of the hill.

In a balanced portfolio, the long bond may variously be slightly better or slightly worse than shorter fixed income in terms of overall efficiency, but far from a slam dunk either way.

Again, what you just said is the truly pernicious problem with the bond scare talk. A poster comes here, hears that they you shouldn't use bonds (or bond funds! a perrenial favorite), they should stick to bank accounts, and upon realizing that they can't use any with the bulk of their money they are told or conclude that oh well, they should just use stocks then, bonds are too scary, the conventional wisdom turned on its head because of one or two percentage points (nevermind that stock themselves are priced to return much less than at other times). If we convince someone who belongs at a 50 or 60% allocation to go to 100%, we've done them a great disservice and put them in danger of massive losses. And we're guilty of thinking, and spreading the notion, that an investor should just look in the market at this or that percentage and change their allocation accordingly because the market is obviously nuts. This is entirely not what we should be doing.

Atgard
Posts: 420
Joined: Wed Apr 09, 2014 2:02 pm

Re: How low are interest rates, really?

Post by Atgard » Tue Apr 29, 2014 12:23 pm

I think you guys may be playing semantic games to argue my core point. Even the "Risk and Time" study you linked proves my point. It is talking about "risk" as a "wide spread of possible outcomes." Yes, the spread between the lowest possible ending dollar amount and highest possible ending dollar amount gets higher over time with stocks. But the odds that the ending value is less than the initial amount invested in real dollars goes down over time. That is the "risk" I stated I am worried about with bonds at these rates. I don't consider the dramatic upside possibility a "risk," and I think you might be torturing the English language if you're saying holding stocks for longer is more "risky" because the ending value could be really, really high.

(That statement could possibly be technically right with enough explanation and rigid definitions and analysis, but is not meaningful to an investor, and will convey an inaccurate impression.)

The graph in the appendix proves my point. It says that over 40 years, stocks have a 10% chance of returning less money than a bank account earning 6% interest. In other words, stocks have a 90% chance to come out ahead of bonds, and a virtual 100% chance to make money (could be a little, could be a whole lot). So, is picking the horse with 90% odds to win more risky? Now, plug in today's 3.4% interest rate and run the numbers again. What are the odds the stock portfolio will have a lower ending balance than the bond portfolio? 0.1%? 0.001%? 0.000%? Unmeasurable, because it has never happened in history?

I will state it again, unequivocally: the odds that a 30- or 40-year bond investment at 3.46% will have a higher ending value than stocks (VTI) over the same time period is infinitesimal. Well under 1%.

If the argument is that the returns are less certain (meaning the standard deviation is much higher) than with a fixed income portfolio, well, that is the definition of "fixed" income, isn't it? But if all of that deviation is to the upside (or even 99.99%), then that is a "risk" I am more than willing to take.

Atgard
Posts: 420
Joined: Wed Apr 09, 2014 2:02 pm

Re: How low are interest rates, really?

Post by Atgard » Tue Apr 29, 2014 12:36 pm

ogd wrote:Again, what you just said is the truly pernicious problem with the bond scare talk. A poster comes here, hears that they you shouldn't use bonds (or bond funds! a perrenial favorite), they should stick to bank accounts, and upon realizing that they can't use any with the bulk of their money they are told or conclude that oh well, they should just use stocks then, bonds are too scary, the conventional wisdom turned on its head because of one or two percentage points (nevermind that stock themselves are priced to return much less than at other times). If we convince someone who belongs at a 50 or 60% allocation to go to 100%, we've done them a great disservice and put them in danger of massive losses. And we're guilty of thinking, and spreading the notion, that an investor should just look in the market at this or that percentage and change their allocation accordingly because the market is obviously nuts. This is entirely not what we should be doing.
First of all, I am not advocating for anyone to do anything, I am trying to further my own understanding and make my own investment decisions for myself. And I CERTAINLY am not trying to change anyone else's asset allocation. I am not trying to "scare" anyone away from bonds. I even said I am nowhere near 100% stocks or 0% fixed income. Everyone has different circumstances, risk tolerance, etc.

I also consider asset allocation across broad classes: stocks, real estate, commodities, and fixed income. Within those, I consider 1.) U.S. stocks, international stocks, sector tilts, etc., (2) REITs, physical real estate, (3) gold & precious metals, oil, futures, etc., and (4) bonds, bond funds, I-bonds, CDs, bank accounts, etc. I determine the overall allocation to each bucket based on time horizon and risk tolerance, like a good Boglehead. BUT, within each bucket, I do not have to buy any particular investment I do not like. For example, I do not buy individual stocks, or sector tilts. I did NOT buy a house in L.A. in 2006 when everyone told me I should. I would not buy a 2-year CD at 0.2%. I would not buy an EE bond over an I-bond. And I would not buy treasuries at current rates. (I do buy I-bonds, other CDs, savings accounts, etc.)

And each person will have access to different options within those asset classes (you may live somewhere housing is a better deal, or join a local credit union with a better CD rate, or not have access to anything preferable to treasuries right now). So I'm sorry if it sounds like I'm generalizing to everyone's situation or saying "No one should ever buy treasuries, in any circumstance." That's not my intention.
ogd wrote:Atgard: leaving aside for a moment that I'm not convinced that history is such a strong guarantee, as opposed to the dates and amounts printed on a bond -- like I said, I don't want to be one of the victims of a historical first...
As for this, aren't you relying on the historical fact that the U.S. has never defaulted on bond payments? Seems unfair that I can't rely on history for stocks while you can with bonds. If stocks can go to zero due to some catastrophic event, then government bonds can default too. (In fact, if such an event occurs, I think it is likely.)

User avatar
ogd
Posts: 4875
Joined: Thu Jun 14, 2012 11:43 pm

Re: How low are interest rates, really?

Post by ogd » Tue Apr 29, 2014 12:45 pm

Atgard wrote:I will state it again, unequivocally: the odds that a 30- or 40-year bond investment at 3.46% will have a higher ending value than stocks (VTI) over the same time period is infinitesimal. Well under 1%.
I personally place it at 10 to 20%. It's intangible, but I think there's just that many things that can go wrong with an economy. I'm not fooled by statistics drawn from a limited set of independent 30 year periods, which are wildly different in terms of historical environment to boot.

But wait, there's more. When we're talking about the remaining 30% of a portfolio that's already 70% in stocks, which we are when we're not forgetting what this is all about, I'd put that chance at 30% or so. When we add into the picture the chance of job loss or medical problem forcing a sale at the worst possible time, it goes even higher. You don't have to keep that 30 year bond to maturity for it to do its job.

If you like statistics from the depths of history, there has already been a time where the only way you kept your home and calorie intake was if you had some bonds. Capitalism is not a smooth ride for just about anyone.

As for the obviousness of six sigma events not happening, tell that to Long Term Capital Management and the taxpayers who had to bail it out.
Atgard wrote:And each person will have access to different options within those asset classes (you may live somewhere housing is a better deal, or join a local credit union with a better CD rate, or not have access to anything preferable to treasuries right now). So I'm sorry if it sounds like I'm generalizing to everyone's situation or saying "No one should ever buy treasuries, in any circumstance." That's not my intention.
When you clearly say that even equities are preferable to Treasuries from a safety angle, we have a problem.
Atgard wrote:As for this, aren't you relying on the historical fact that the U.S. has never defaulted on bond payments? Seems unfair that I can't rely on history for stocks while you can with bonds. If stocks can go to zero due to some catastrophic event, then government bonds can default too. (In fact, if such an event occurs, I think it is likely.)
I am not seeing the world in black and white. I acknowledge the chance of default and hoping that if that happens, stocks will bail me out, because I do hold both to balance their various risks. Although chances are that if the US does such a boneheaded thing nothing will keep its value, so for that my hope is people's sanity, and my vote.

Atgard
Posts: 420
Joined: Wed Apr 09, 2014 2:02 pm

Re: How low are interest rates, really?

Post by Atgard » Tue Apr 29, 2014 1:08 pm

ogd wrote:When you clearly say that even equities are preferable to Treasuries from a safety angle, we have a problem.
If you are sure not to touch them for 30+ years -- please don't leave that part of my statement out. In the real world, most people are not sure of that (certainly not with 100% of their money), will panic at the bottom, etc. -- hence why I am not nor do I advocate 100% stocks. (I do not advocate anything to anyone reading this forum.)

Anyway, I'm not trying to change anyone's mind, so I won't continue. You guys can have the last word. I have made up my own mind on what to do for my own financial situation. So thank you for the back-and-forth on that.

thx1138
Posts: 1033
Joined: Fri Jul 12, 2013 2:14 pm

Re: How low are interest rates, really?

Post by thx1138 » Tue Apr 29, 2014 1:29 pm

Atgard wrote:
ogd wrote:When you clearly say that even equities are preferable to Treasuries from a safety angle, we have a problem.
If you are sure not to touch them for 30+ years -- please don't leave that part of my statement out.
This is in fact the whole point to varying asset allocation over time. To reach a given retirement goal you may *need* to collect risk premia. Failing to collect those risk premia can guarantee portfolio failure in many situations. That's a form of lack of safety itself. In that sense equities can be thought of as safer than treasuries. Jumping off the railroad trestle into the water as a train approaches does risk drowning and broken limbs, but remaining where you are standing is the riskier option despite the fact there is no volatility in the outcome if you choose to do so :shock:

Risk and safety are not just down to volatility. For most all practical retirement investing it is the floor that matters, not the volatility above it. Over long holding periods equities provide a higher floor than the risk free rate (treasuries) hence the reason you have very high equity allocations when far from retirement.

However, one also has to be careful not to fall into the highest expected return trap either. Again, it is about the floor and not the average or the volatility. Historically portfolio success is a bit better at allocations below 100/0 and so that again suggests completely eschewing fixed income isn't the optimal answer either.

User avatar
Phineas J. Whoopee
Posts: 9169
Joined: Sun Dec 18, 2011 6:18 pm

Re: How low are interest rates, really?

Post by Phineas J. Whoopee » Tue Apr 29, 2014 2:06 pm

ogd wrote:...
As for the obviousness of six sigma events not happening, tell that to Long Term Capital Management and the taxpayers who had to bail it out.
...
I don't like contradicting you, ogd, because usually you and I agree, however:

A) Taxpayers did not bail out LTCM. Alan Greenspan, the then green Fed chair, used moral suasion to organize a private counterparty rescue, on the grounds that not doing so would be still worse for the rescuers. While the financial system was at some risk of disruption, although not actual collapse, public monies weren't used. LTCM in fact failed. Its creditors and service providers, and even some others, along with LTCM's investors, bore the loss.

B) It was not a six sigma event. The LTCM managers only claimed it was to absolve themselves of responsibility. Six sigma might have been what their computer models they themselves didn't understand said. When models and the real world disagree, it's the real world which wins. In theory, of course, there is no difference between theory and practice.

PJW

User avatar
ogd
Posts: 4875
Joined: Thu Jun 14, 2012 11:43 pm

Re: How low are interest rates, really?

Post by ogd » Tue Apr 29, 2014 2:28 pm

Atgard wrote:If you are sure not to touch them for 30+ years -- please don't leave that part of my statement out. In the real world, most people are not sure of that (certainly not with 100% of their money), will panic at the bottom, etc. -- hence why I am not nor do I advocate 100% stocks. (I do not advocate anything to anyone reading this forum.)
That's right, the only way you can be sure not to touch even the "first" 20% or 40% of your portfolio is if you're a thought experiment, not a real person with a job and a body that can both fail.

I think your big mistake in the argument about the long bond is assuming it moves the goalpost to 30 years and you necessarily care only about that term. This is not true. It's a very liquid investment with very clear price behavior -- by selling at any given time within those 30 years you forsake a certain amount of future money, roughly equal to the remaining term times the then-prevailing market rates. So now you look at the behavior of market rates in recessions and you realize that they are generally not affected and often counter-affected by the general havoc around them, which makes the 30 year a very good thing to sit on in such situations. What you have in the context of a portfolio is an investment paying 3.5% throughout that period, whose price moves in a certain, sometimes desirable way in response to market conditions, and research shows that the longer term enhances the efficiency of an equity-heavy portfolio rather than hurting it through volatility, even for the past data that I argue is equity-happy to begin with. The 3.5% long bond is not a guaranteed loser in this context, far from it.

PJW: thanks for the correction! It's still the case that parties arguably not at fault had to bear the cost, in preference to one even greater. As for six sigma, no event is going to look like a six sigma after the fact because such things simply don't happen, it's just that bad or over-optimistic models can convince you you don't need to worry about something that you definitely need to worry about. Such as poor equity returns for long periods, which seems so clearly possible to me that I can't understand how anyone can argue it's as infinitesimal as an asteroid hit or something. It's not.

Atgard
Posts: 420
Joined: Wed Apr 09, 2014 2:02 pm

Re: How low are interest rates, really?

Post by Atgard » Thu May 01, 2014 6:39 pm

An honest question (I don't want to debate the merit of bonds in general or suggest anyone change their asset allocation)… but you say the 30-year bond is liquid and doesn't need to be held to term, which is of course true. But then aren't you are talking about a more risky investment, since if you sell before maturity the price can go up or down (and will go down if interest rates rise)? I mean, a big draw of bonds is the "guarantee" you speak about, but the guaranteed return of principal + interest only holds true if you hold to maturity. If not, the value could go down and someone could panic and sell at a loss, just like with stocks, right? Or am I missing something?

Besides, we were having a mostly-theoretical discussion about where to put money you were pretty sure you wouldn't need for 30 years. And no one should put ALL their money in such an investment (stocks or bonds), unless they're going on a long trip to Mars or something. :P

User avatar
ogd
Posts: 4875
Joined: Thu Jun 14, 2012 11:43 pm

Re: How low are interest rates, really?

Post by ogd » Thu May 01, 2014 8:02 pm

Atgard wrote:But then aren't you are talking about a more risky investment, since if you sell before maturity the price can go up or down (and will go down if interest rates rise)? I mean, a big draw of bonds is the "guarantee" you speak about, but the guaranteed return of principal + interest only holds true if you hold to maturity. If not, the value could go down and someone could panic and sell at a loss, just like with stocks, right? Or am I missing something?
A safe bond has the same risk whether held to maturity or not, since the losses are the same either way (opportunity losses in the first case). Or, if you prefer, the chance and magnitude of underperforming cash is the same, which is something I call a risk regardless where you put the goalpost for measuring a standard deviation. The differences are illusory.

Here's where the guarantee comes in: if my bonds decline in value, I know for a fact that they will go back up within the term. It's written in black and white. Nothing of the sort exists for stocks, and were you to buy it from some option writer it would be so expensive that it would likely kill any extra returns. Interest rate is a different kind of risk, one with a guaranteed rebound. Furthermore, the market knows this, which is why the bond prices have such desirable behavior in recessions. There's no question that the longer the bond the more volatile, but it's uncorrelated volatility and it does come with rewards.

Now for panic selling: it's true that you can do it with bonds and hurt yourself. But there are two mitigating factors: 1) the losses are much smaller than stocks to begin with, and 2) the math of bonds is such that what you lose going forward is an income stream equal to the prevailing rates times the remaining term. Which, given that rates have gone up so much, should be easily replaceable. For example, even cash might do the same job as the bond you just sold.
Argard wrote:Besides, we were having a mostly-theoretical discussion about where to put money you were pretty sure you wouldn't need for 30 years. And no one should put ALL their money in such an investment (stocks or bonds), unless they're going on a long trip to Mars or something. :P
The gist of the argument as I remember it was whether anyone has a reason to buy current 30 year bonds. Neither of us (looks like) is talking about an investment in isolation. There are clear reasons to hold these bonds in a portfolio.

User avatar
ogd
Posts: 4875
Joined: Thu Jun 14, 2012 11:43 pm

Re: How low are interest rates, really?

Post by ogd » Thu May 01, 2014 8:02 pm

Atgard wrote:But then aren't you are talking about a more risky investment, since if you sell before maturity the price can go up or down (and will go down if interest rates rise)? I mean, a big draw of bonds is the "guarantee" you speak about, but the guaranteed return of principal + interest only holds true if you hold to maturity. If not, the value could go down and someone could panic and sell at a loss, just like with stocks, right? Or am I missing something?
A safe bond has the same risk whether held to maturity or not, since the losses are the same either way (opportunity losses in the first case). Or, if you prefer, the chance and magnitude of underperforming cash is the same, which is something I call a risk regardless where you put the goalpost for measuring a standard deviation. The differences are illusory.

Here's where the guarantee comes in: if my bonds decline in value, I know for a fact that they will go back up within the term. It's written in black and white. Nothing of the sort exists for stocks, and were you to buy it from some option writer it would be so expensive that it would likely kill any extra returns. Interest rate is a different kind of risk, one with a guaranteed rebound. Furthermore, the market knows this, which is why the bond prices have such desirable behavior in recessions. There's no question that the longer the bond the more volatile, but it's uncorrelated volatility and it does come with rewards.

Now for panic selling: it's true that you can do it with bonds and hurt yourself. But there are two mitigating factors: 1) the losses are much smaller than stocks to begin with, and 2) the math of bonds is such that what you lose going forward is an income stream equal to the prevailing rates times the remaining term. Which, given that rates have gone up so much, should be easily replaceable. For example, even cash might do the same job as the bond you just sold.
Argard wrote:Besides, we were having a mostly-theoretical discussion about where to put money you were pretty sure you wouldn't need for 30 years. And no one should put ALL their money in such an investment (stocks or bonds), unless they're going on a long trip to Mars or something. :P
The gist of the argument as I remember it was whether anyone has a reason to buy current 30 year bonds. Neither of us (looks like) is talking about an investment in isolation. There are clear reasons to hold these bonds in a portfolio.

Post Reply