Static vs Dynamic Asset Allocation; Victor Meets the Boglehead

[watchnerd]

It flies in the face of so much of the default retail financial planning thinking.

The Trinity study, the so-called "4% rule", endless backtesting SWR financial calculation tools, static AA, all fall outside the framing.

It requires people to basically set aside their existing mental models and think from a totally different perspective.

Honestly, it makes a lot of intuitive sense to me when I frame it in terms of (i) "I want to allocate my portfolio such that equities give me x percent of the portfolio risk and bonds give me (100 - x) percent of the risk" or something similar, and (ii) "I am willing to update my portfolio so that risk allocation stays approximately true". You define the time frame that you are willing to update on (weeks, months, years) and match the parameters to that time frame.

A fixed 80/20 portfolio allocation falls within that framework if you are working with very long lookback periods for the parameters.

I could see that one might use an intermediate time frame (5-10 years) as a way of gliding through the life cycle stages.

That's how I think about it, too.

Managing the risk / reward budget is the intent, not trying to beat the market.

[IDpilot]

Yes, you use historical data. In The Missing Billionaires on page 53 the authors explain it this way, "What we can do is build up an estimate for market risk using a blend of long- and short-horizon realized volatility" and then refer you via a footnote to an article on their website. In the appendix of that article they give this explanation;

We calculate a series of rolling 10-year equity volatility and rolling 2-year equity volatility from monthly closing prices of the S&P500. We then take a weighted average of the life-to-date average of the rolling 10-year volatility and the most recent 2-year volatility. We put 75% and 25% weight on the 10-year and 2-year volatility measures, both expressed as variances, and then take the square root of that weighted average to arrive at the spot estimate of equity volatility that an investor might reasonably have used in deciding how much equity exposure to take using the Merton Rule. The chart below shows the volatility estimate we used in the historical simulation.

https://elmwealth.com/earnings-yield-dy ... llocation/

This is the major issue I have with Elm. Lots of seemingly arbitrary approaches, assumptions, caps, etc. on how they adjust AA -- especially on the volatility / momentum side.

This approach is the idea behind GARCH models and is an industry standard tool for modeling financial time series that exhibit time-varying volatility and volatility clustering.

[IDpilot]

More or less. In the first place, short-term volatility usually doesn't rise during periods of high realized returns, generally it falls. Volatility itself, as well as implied volatility (Black-Scholes) generally rises during periods of highly negative realized returns.

One can do one of three things when that happens: nothing--ie, not rebalance; rebalance back to policy; or "overbalance," ie, increase the policy allocation. All three of those things are reasonable. What's *not* reasonable is to use the Merton Share formula, which is to lower one's allocation, which would mean selling an asset after price falls.

Is the following correct? Haghani and White use a historical volatility input rather than a short term number. Therefore, when equity prices drop (at least in part) due to a change in valuations (declining PE ratio), Merton share would indicate increasing equity percentage. Short term volatility spikes would not be reflected. This would lead to buy low, not sell low, right?

I think this would also be true for TPAW, since it uses historical volatility inputs.

Thanks!

Haghani and White emphasis the most recent 2-year volatility. In their own words here is what they do,

We calculate a series of rolling 10-year equity volatility and rolling 2-year equity volatility from monthly closing prices of the S&P500. We then take a weighted average of the life-to-date average of the rolling 10-year volatility and the most recent 2-year volatility. We put 75% and 25% weight on the 10-year and 2-year volatility measures, both expressed as variances, and then take the square root of that weighted average to arrive at the spot estimate of equity volatility...

Yes, TPAW uses the historical volatility. Currently that is 18%.

As to what happens when equity prices drop due to changing valuations the answer is "it depends" since you can't address that scenario without filling in a few more blanks like what happened to the risk-free rate of return and volatility. Remember that the Merton Share is the expected excess return of risky asset above safe asset return divided by the product of coefficient of risk-aversion and the square of the variability of the return of the risky asset.

[Circle the Wagons]

This is the major issue I have with Elm. Lots of seemingly arbitrary approaches, assumptions, caps, etc. on how they adjust AA -- especially on the volatility / momentum side.

This approach is the idea behind GARCH models and is an industry standard tool for modeling financial time series that exhibit time-varying volatility and volatility clustering.

They simplify all that into a low / neutral / high risk overlay.
They cap movements off of baseline (even at full scaling).
They ignore duration.
Etc.

[IDpilot]

This approach is the idea behind GARCH models and is an industry standard tool for modeling financial time series that exhibit time-varying volatility and volatility clustering.

They simplify all that into a low / neutral / high risk overlay.
They cap movements off of baseline (even at full scaling).
They ignore duration.
Etc.

I get that you don't like much/anything of what Haghani and White do. But to characterize how they estimate volatility as "seemingly arbitrary" seems a bit off base.

[Circle the Wagons]

They simplify all that into a low / neutral / high risk overlay.
They cap movements off of baseline (even at full scaling).
They ignore duration.
Etc.

I get that you don't like much/anything of what Haghani and White do. But to characterize how they estimate volatility as "seemingly arbitrary" seems a bit off base.

I like a lot of what they do. And I like the book.

But the caps, for instance, are an example of the seemingly arbitrary. +/- two-thirds on valuation and +/- one-third on volatility, I believe.

In my post above, I should have clarified that these are my major issues within a sea of things I mostly like -- like their use of the ERP, low expense ratio, transparency of approach, walking the talk with their own money ...

[comeinvest]

I do not understand how to measure expected volatility - do you use historical data? from the elm's website, it seems more like some momentum factor instead.

another thing i'm not so sure is the time: since we are using 10-year TIPS, we should measure 10 year volatility, not daily, correct? 10-year TIPS, holding it will gradually reduce the duration until 0. The stocks are kind of like a rolling 10 year bond fund, where the duration is "constant". How is this comparable?

Yes, you use historical data. In The Missing Billionaires on page 53 the authors explain it this way, "What we can do is build up an estimate for market risk using a blend of long- and short-horizon realized volatility" and then refer you via a footnote to an article on their website. In the appendix of that article they give this explanation;

We calculate a series of rolling 10-year equity volatility and rolling 2-year equity volatility from monthly closing prices of the S&P500. We then take a weighted average of the life-to-date average of the rolling 10-year volatility and the most recent 2-year volatility. We put 75% and 25% weight on the 10-year and 2-year volatility measures, both expressed as variances, and then take the square root of that weighted average to arrive at the spot estimate of equity volatility that an investor might reasonably have used in deciding how much equity exposure to take using the Merton Rule. The chart below shows the volatility estimate we used in the historical simulation.

https://elmwealth.com/earnings-yield-dy ... llocation/

Based on chart 7 on https://elmwealth.com/earnings-yield-dy ... llocation/ , it looks like their 2-year and 10-year volatility mix was in a narrow range in the ca. 80 years post-war, with a few short spikes that were probably hard to anticipate based on lookback. So the varying real risk-free rates and varying equity valuations were probably the main ingredients to the varying AA.

Thinking about it, from a conceptual asset allocation point of view, shouldn't we compare the excess returns of both asset classes, equities and TIPS, over the short-term risk-free rates for purposes of asset allocation? I'm still a bit confused here. In a leveraged asset allocation, the excess return over risk-free (i.e. over the cost of leverage) is the actual return of the respective position to the portfolio - what is called the term premium for bonds. In an unleveraged portfolio, the same principle applies if you measure the returns vs. the risk-free cash that you could alternatively invest in. (Could take the real risk-free rate, which has zero drawdown and volatility.) Portfolio metrics like Sharpe ratio are typically measured by returns against the risk-free rate.

[comeinvest]

Honestly, it makes a lot of intuitive sense to me when I frame it in terms of (i) "I want to allocate my portfolio such that equities give me x percent of the portfolio risk and bonds give me (100 - x) percent of the risk" or something similar, and (ii) "I am willing to update my portfolio so that risk allocation stays approximately true". You define the time frame that you are willing to update on (weeks, months, years) and match the parameters to that time frame.

A fixed 80/20 portfolio allocation falls within that framework if you are working with very long lookback periods for the parameters.

I could see that one might use an intermediate time frame (5-10 years) as a way of gliding through the life cycle stages.

That's how I think about it, too.

Managing the risk / reward budget is the intent, not trying to beat the market.

It's the same thing though. One leads to the other.
Give me an asset or an AA with higher combined Sharpe ratio, and I'll most likely beat you over a lifetime, at much lower shortfall risk.

[comeinvest]

Is the following correct? Haghani and White use a historical volatility input rather than a short term number. Therefore, when equity prices drop (at least in part) due to a change in valuations (declining PE ratio), Merton share would indicate increasing equity percentage. Short term volatility spikes would not be reflected. This would lead to buy low, not sell low, right?

I think this would also be true for TPAW, since it uses historical volatility inputs.

Thanks!

Personally, I prefer to pick a volatility number and stick with it, not try to make it a moving target as the fact that it's squared in the Merton equation means small changes lead to big swings in the denominator.

It's the change in the numerator that I want to vary with stock valuation and pay attention to.

Per user Hydromod, you would then basically forfeit the benefit of using volatility clustering as part of your forecast model. Per the evidence that he presented, this would mean leaving money on the table.

[watchnerd]

Personally, I prefer to pick a volatility number and stick with it, not try to make it a moving target as the fact that it's squared in the Merton equation means small changes lead to big swings in the denominator.

It's the change in the numerator that I want to vary with stock valuation and pay attention to.

Per user Hydromod, you would then basically forfeit the benefit of using volatility clustering as part of your forecast model. Per the evidence that he presented, this would mean leaving money on the table.

*shrug*

Okay, so some money gets theoretically left on the table.

It's a model, not a precise forecasting tool.

[ScubaHogg]

Thinking about it, from a conceptual asset allocation point of view, shouldn't we compare the excess returns of both asset classes, equities and TIPS, over the short-term risk-free rates for purposes of asset allocation?

I think I understand the question. And if I do I believe it’s because of their use of the expected utility framework. It’s much more focused on total lifetime consumption than maximizing total lifetime returns. Duration matched TIPS provide the closet thing we have to a risk-free income stream. Looking at it that way there’s nothing risk free about short term nominal rates

So it’s a question, “how much future utility do I want to risk in exchange for the higher expected return of equities.”

[alluringreality]

Thinking about it, from a conceptual asset allocation point of view, shouldn't we compare the excess returns of both asset classes, equities and TIPS, over the short-term risk-free rates for purposes of asset allocation?

I think I understand the question. And if I do I believe it’s because of their use of the expected utility framework. It’s much more focused on total lifetime consumption than maximizing total lifetime returns. Duration matched TIPS provide the closet thing we have to a risk-free income stream. Looking at it that way there’s nothing risk free about short term nominal rates

So it’s a question, “how much future utility do I want to risk in exchange for the higher expected return of equities.”

Personally I tend to define Series I Savings Bonds as generally the least risky asset that I can purchase, although purchase limits tend to complicate how large a percentage of Bogleheads assets they can easily cover. I bonds are non-marketable, so there's zero nominal near-term risk after the first year compared to TIPS, although I bond real rates are delayed as nominal payments, which results in some near-term risk in real terms. I'd consider it plausible that a model might allocate between I bonds (cash-like after the first year), TIPS (bonds), and stocks. The minimum fixed rate for I bonds is 0% real, and TIPS or stocks could potentially offer lower projected returns. Generally I resist guessing if the Fed's flexible average inflation targeting might end up affecting longer-term inflation perceptions, or near-term nominal rate expectations, but Cullen Roche's comments still appear solidly on the side of no change in outlook.
viewtopic.php?p=7484980#p7484980

[prioritarian]

(*because we have a TIPS LMP ladder, we exclude Treasuries from the risk port, otherwise we'd be massively overweight Treasuries)

If your TREASURY bond LMP* massively outewighs your "risk port" then it is not a risk portfolio at all -- it's a very conservative aggregate portfolio with a low expected return. IMO, the bucketing approaches that are so popular here are a mild form of cognitive dissonance.


* inflation-protected treasuries have thus far performed similarly to treasuries (across duration)

[watchnerd]

(*because we have a TIPS LMP ladder, we exclude Treasuries from the risk port, otherwise we'd be massively overweight Treasuries)

If your TREASURY bond LMP* massively outewighs your "risk port" then it is not a risk portfolio at all -- it's a very conservative aggregate portfolio with a low expected return. IMO, the bucketing approaches that are so popular here are a mild form of cognitive dissonance.


* inflation-protected treasuries have thus far performed similarly to treasuries (across duration)

You seem to have misunderstood the point.

Let's say my aggregate portfolio is 50% risk port and 50% TIPS.

If the risk portfolio is a global market weight portfolio of stocks and bonds, the risk portfolio bond portion will already contain nominal Treasuries as the dominant bond allocation by market weight.

Combined with the 50% in TIPS, that means my total exposure to US Treasuries would be waaaaay above market weight.

So I exclude nominal Treasuries from the risk portfolio entirely, leading to the risk port AA in my signature.

In the 50/50 example above, if we lump the alts in with stocks as 'risky' assets, we get 83% risk assets * 50% risk port = 41.5% risk assets (mostly stocks) for the aggregate portfolio weighting.

40% stocks is a conservative balanced allocation, for sure, but I don't think I'd go as far as calling it low return and very conservative.

For comparison, the Vanguard Target Retirement Date 2025 fund is about 50% stock.

[prioritarian]

Combined with the 50% in TIPS, that means my total exposure to US Treasuries would be waaaaay above market weight.

This makes complete sense.

[comeinvest]

Thinking about it, from a conceptual asset allocation point of view, shouldn't we compare the excess returns of both asset classes, equities and TIPS, over the short-term risk-free rates for purposes of asset allocation?

I think I understand the question. And if I do I believe it’s because of their use of the expected utility framework. It’s much more focused on total lifetime consumption than maximizing total lifetime returns. Duration matched TIPS provide the closet thing we have to a risk-free income stream. Looking at it that way there’s nothing risk free about short term nominal rates

So it’s a question, “how much future utility do I want to risk in exchange for the higher expected return of equities.”

I get your point of view, which is perhaps the dominant point of view in this thread regarding the "risk-free" part of the asset allocation; but I have been mostly following and contributing to the leveraged lifetime (also called lifecycle) asset allocation threads (HFEA, mHFEA), which consider multi-asset portfolios where the premise is that both equities and treasuries (can conceptually replace treasuries with TIPS for the discussion at hand) are risky components of the combined portfolio, and the combined asset allocation which can vary over the lifetime (time diversification) seeks to optimize the combined risk/return given expected returns, volatilities/risk, and expected correlations. This is obviously detached from the concept of "naive" liability matching in the sense of trying to find a risk-free asset that duration-matches a specific individual investor's liabilities, against which another ("risky") asset is measured against. I say "naive" (no offense intended), because the overarching optimization problem should be to increase overall (probabilistic) utility, which generally means maximizing the percentile range of "final" lifetime investing outcomes while minimizing shortfall risk, which is arguably more likely accomplished with the "leverage unconstrained" multi-asset lifetime AA models with time diversification.
In this context, like I said, none of the component assets are "risk-free"; in fact they are often on purpose leveraged to similar volatilities, which should result in better diversification of overall risk exposure. Overall risk exposure would be a combination of exposure to the equity risk premium and exposure to the term premium. Term premium is the performance premium of treasuries (or OIS swaps, if treasuries are no longer risk-free) relative to cash. (Total risk can still be dialed at will by choosing whatever leverage factors either <1 or >1.)
I subscribed to this thread because I want to explore the possibility of marrying the concept of dynamic AA with the concept of leveraged multi-asset portfolios, in particular leveraged stock/bond portfolios (e.g. mHFEA). In this scenario, I guess both stocks (equities) and bonds (treasuries and/or TIPS) can independently have time-varying valuations, expected returns and risk/volatilities, that might be inputs to the dynamic AA of both components; think of ZIRP (period of zero interest rate policy and low long-maturity treasury yields) when some advocated eliminating the treasuries position from mHFEA due to expected eventual mean reversion and/or low expected term premia and high risk (while others advocated staying the course with a static AA). I am hoping that within this "multivariate" framework or context, a solution to the dynamic AA puzzle would cover the more specific scenarios discussed so far as a special case.
I hope this clarifies where I was coming from when I asked the question.

[Ben Mathew]

Duration matched TIPS provide the closet thing we have to a risk-free income stream. Looking at it that way there’s nothing risk free about short term nominal rates

Yes. Some perspective on this from Professor John Cochrane in Portfolios for Long-Term Investors (Pages 4-5):

• 2. The Payoff Perspective
  
  I am inspired by the brilliant Campbell and Viceira (2001) article on long-term bonds. The indexed perpetuity is the riskless asset for a long-run investor. Stop and savor that statement. I remember first seeing the paper when John gave it at Chicago. I instantly thought, “This is obvious.” And then, I realized that neither I nor anyone in the room knew it, and I realized “this is brilliant.”
  
  Now, the statement is only obvious if you look at the payoffs. An indexed perpetuity gives a steady stream of income forever. It’s the risk-free payoff. Duh. The statement is not at all obvious if you look at it Merton-style: When bond prices fall, bond expected returns rise. The indexed perpetuity’s expected return happens to rise just enough when its price falls, that the dynamically-optimizing consumer can construct a riskless consumption stream. Seeing that fact takes a lot of dynamic programming.
  
  This verity is not at all obvious to the investing world. The idea that indexed perpetuities are the riskfree asset meets incredulity. How many MBA investment classes yet mention this obvious fact, instead calling money market funds the “riskless asset?”
  

[Northern Flicker]

The inflation-adjusted perpetuity is only a risk-free asset if that return schedule matches your liabilities. I could live to 100 with uniform real liabilities up to the time of death. Or I could need 5 years of skilled nursing care starting at age 70 and die at 75. If we define risk as the risk of not bring able to cover our expenses, then the fact that either outcome could materialize shows that the perpetuity is not risk-free. If rates rise sharply until I turn 70, and then need to accelerate the rate of spending greatly, the perpetuity will have lost substantial value just when I need to withdraw from it at a high rate.

Liability matching is the bread and butter of the life insurance industry. They don't need perpetuities because they are matching to the average liabilities of a large actuarial pool. Retirees are an actuarial pool of size 1 or 2 (single or married). Our outcomes have much higher variance than the average outcome of a large actuarial pool. Any outcome we match to takes the risk of experiencing a different outcome.

A real perpetuity or SPIA can be risk free if it is funding uniformly distributed liabilities, such as an income floor for predictable expenses. Overfunded retirees may have the luxury of redundantly matching to different outcomes.

[IDpilot]

I get that you don't like much/anything of what Haghani and White do. But to characterize how they estimate volatility as "seemingly arbitrary" seems a bit off base.

I like a lot of what they do. And I like the book.

But the caps, for instance, are an example of the seemingly arbitrary. +/- two-thirds on valuation and +/- one-third on volatility, I believe.

In my post above, I should have clarified that these are my major issues within a sea of things I mostly like -- like their use of the ERP, low expense ratio, transparency of approach, walking the talk with their own money ...

Where are you seeing these caps? I'm confused.

[IDpilot]

If your TREASURY bond LMP* massively outewighs your "risk port" then it is not a risk portfolio at all -- it's a very conservative aggregate portfolio with a low expected return. IMO, the bucketing approaches that are so popular here are a mild form of cognitive dissonance.


* inflation-protected treasuries have thus far performed similarly to treasuries (across duration)

You seem to have misunderstood the point.

Let's say my aggregate portfolio is 50% risk port and 50% TIPS.

If the risk portfolio is a global market weight portfolio of stocks and bonds, the risk portfolio bond portion will already contain nominal Treasuries as the dominant bond allocation by market weight.

Combined with the 50% in TIPS, that means my total exposure to US Treasuries would be waaaaay above market weight.

So I exclude nominal Treasuries from the risk portfolio entirely, leading to the risk port AA in my signature.

In the 50/50 example above, if we lump the alts in with stocks as 'risky' assets, we get 83% risk assets * 50% risk port = 41.5% risk assets (mostly stocks) for the aggregate portfolio weighting.

40% stocks is a conservative balanced allocation, for sure, but I don't think I'd go as far as calling it low return and very conservative.

For comparison, the Vanguard Target Retirement Date 2025 fund is about 50% stock.

Which is it? The risk portfolio can't be a "global market weight portfolio of stocks and bonds" and "exclude nominal Treasuries."

[ScubaHogg]

The inflation-adjusted perpetuity is only a risk-free asset if that return schedule matches your liabilities. I could live to 100 with uniform real liabilities up to the time of death. Or I could need 5 years of skilled nursing care starting at age 70 and die at 75. If we define risk as the risk of not bring able to cover our expenses, then the fact that either outcome could materialize shows that the perpetuity is not risk-free.

It’s “riskless” in the sense that the payoff is risk-less, as compared to an uncertain payoff. Not in the sense that it’ll somehow make your future expenses known with certainty. Even a sufficiently large perpetuity would solve the problem you outline. Solving your problem is a matter of degree, not kind.

In a broader sense I don’t really understand your complaint about lumpy expenses. Folks always throw this out as a slam dunk against TIPS ladders, LMP, etc. But I’ve never seen, not a single time, someone say their income from working is somehow risky because their expenses are sometimes lumpy.

[KEotSK66]

Who's going to claim a perpetuity when it doesn't meet their liabilities/needs???

My draw, personal inflation expectation, and portfolio are designed to be a perpetuity in that my portfolio on the day I retired will grow with personal inflation AND meet my monthly needs adjusted for personal inflation.

I'm a little bit behind on my required annualized TR to achieve the above but I expect to be back on track in the coming years, the early 2020s weren't so good for me.

[alluringreality]

In a broader sense I don’t really understand your complaint about lumpy expenses. Folks always throw this out as a slam dunk against TIPS ladders, LMP, etc. But I’ve never seen, not a single time, someone say their income from working is somehow risky because their expenses are sometimes lumpy.

Are the two situations parallels? Often it's possible to borrow against an income stream from working. The collapse of Silicon Valley Bank comes to mind for an example of asset duration potentially exceeding an elastic investment horizon, liquidity constraints, or systematically undervaluing liquidity. I'm not exactly following why working and asset selection amount to consistent examples.

[Circle the Wagons]

I like a lot of what they do. And I like the book.

But the caps, for instance, are an example of the seemingly arbitrary. +/- two-thirds on valuation and +/- one-third on volatility, I believe.

In my post above, I should have clarified that these are my major issues within a sea of things I mostly like -- like their use of the ERP, low expense ratio, transparency of approach, walking the talk with their own money ...

Where are you seeing these caps? I'm confused.

https://elmwealth.com/our-asset-allocation-methodology/

[ScubaHogg]

In a broader sense I don’t really understand your complaint about lumpy expenses. Folks always throw this out as a slam dunk against TIPS ladders, LMP, etc. But I’ve never seen, not a single time, someone say their income from working is somehow risky because their expenses are sometimes lumpy.

Are the two situations parallels? Often it's possible to borrow against an income stream from working. The collapse of Silicon Valley Bank comes to mind for an example of asset duration potentially exceeding an elastic investment horizon, liquidity constraints, or systematically undervaluing liquidity. I'm not exactly following why working and asset selection amount to consistent examples.

Off hand you can,

A) margin loans against a portfolio
B) home equity loans
C) even a tips ladder has a giant pile of marketable assets. If you sell them early the return won’t be perfectly known in advance, but how much are we talking about? Selling a TIP one year early won’t incur that much interest rate risk
D) if you go into a nursing home and die at 75, you might have big lumpy expenses. But if you’ve got a tips ladder to 100 you’ve got 25 years worth of tips available for liquidation

The “lumpy expenses” complaint is just grasping at straws for a criticism of an LMP

[IDpilot]

Where are you seeing these caps? I'm confused.

https://elmwealth.com/our-asset-allocation-methodology/

I see. None of the caps mention in that 2015 article are mentioned in Missing Billionaires.

[Northern Flicker]

It’s “riskless” in the sense that the payoff is risk-less, as compared to an uncertain payoff. Not in the sense that it’ll somehow make your future expenses known with certainty.

It is all about the risk model used. If your model of risk is variance of real income generated over indefinite time, then a real perpetuity minimizes it.

If I define risk as the probability of running out of money, the real perpetuity certainly is far from risk-free, and does not even minimize risk.

Even a sufficiently large perpetuity would solve the problem you outline. Solving your problem is a matter of degree, not kind.

True-- if assets are 300x expenses, and the risk model is the probability of running out of money, then a fairly wide range of portfolios are virtually risk-free. One could argue that a hypothetical large real perpetuity generating income that is 10x expenses would protect against hyperinflation if the inflation adjustment could be trusted to stand up to that.

In a broader sense I don’t really understand your complaint about lumpy expenses. Folks always throw this out as a slam dunk against TIPS ladders, LMP, etc.

LMP's neutralize risk by matching to liabilities. If the liabilities are lumpy, the LMP payout needs to be lumpy if the risk is to be neutralized.

But I’ve never seen, not a single time, someone say their income from working is somehow risky because their expenses are sometimes lumpy.

Seriously? Lots of working people incur lumpy expenses beyond the ability of their income to meet. That's why loans and financing exist. That's why people prefer fixed term mortgages, so that lumpy house payments don't lead to losing their home. It also is why insurance products exist-- an entire industry whose business is turning very lumpy liabilities into smoother ones. Countries also smooth out lumpy liabilities to varying degrees by socializing risk and liabilities to varying degrees.

That was not the analogy to use to try to make your point.

[ScubaHogg]

Seriously? Lots of working people incur lumpy expenses beyond the ability of their income to meet. That's why loans and financing exist. That's why people prefer fixed term mortgages, so that lumpy house payments don't lead to losing their home. It also is why insurance products exist-- an entire industry whose business is turning very lumpy liabilities into smoother ones. Countries also smooth out lumpy liabilities to varying degrees by socializing risk and liabilities to varying degrees.

That was not the analogy to use to try to make your point.

I never said people don't have lumpy expenses. I said no one calls wage income risky because expenses are lumpy. They do call TIPS ladders risky because expenses are lumpy...for reasons.

It's only a criteria that's applied to post-retirement income streams

And anyone with financial assets has plenty of financing access. Home equity loans, reverse mortgages, credit cards, consumer loans, margin loans, risk portfolio realizations, and so on

[ScubaHogg]

True-- if assets are 300x expenses, and the risk model is the probability of running out of money, then a fairly wide range of portfolios are virtually risk-free.

That's silly. Even a 2x perpetuity over minimum living expenses would cover most anything the vast majority of people will actually experience*

*assuming one isn't one of those Bogleheads who wants to plan for 25 years of memory care plus 24/7/365 extra caregivers plus a large negative legal judgement plus dinosaurs coming back to life, etc

[Northern Flicker]

Seriously? Lots of working people incur lumpy expenses beyond the ability of their income to meet. That's why loans and financing exist. That's why people prefer fixed term mortgages, so that lumpy house payments don't lead to losing their home. It also is why insurance products exist-- an entire industry whose business is turning very lumpy liabilities into smoother ones. Countries also smooth out lumpy liabilities to varying degrees by socializing risk and liabilities to varying degrees.

That was not the analogy to use to try to make your point.

I never said people don't have lumpy expenses. I said no one calls wage income risky because expenses are lumpy. They do call TIPS ladders risky because expenses are lumpy...for reasons.

Actually, people do talk about the need to have some savings to cover lumpy expenses that a wage income stream won't cover. I've never seen anyone call a TIPS ladder risky.

[ScubaHogg]

Actually, people do talk about the need to have some savings to cover lumpy expenses that a wage income stream won't cover.

Well I agree people should have some liquidity to meet lumpy expenses

I've never seen anyone call a TIPS ladder risky.

Are you sure?

Northern Flicker wrote: ↑Sun Dec 01, 2024 12:51 am
The inflation-adjusted perpetuity is only a risk-free asset if that return schedule matches your liabilities. I could live to 100 with uniform real liabilities up to the time of death. Or I could need 5 years of skilled nursing care starting at age 70 and die at 75. If we define risk as the risk of not bring able to cover our expenses, then the fact that either outcome could materialize shows that the perpetuity is not risk-free

Since a TIPS ladder is the closest thing we can purchase to an inflation indexed perpetuity…

[alluringreality]

Reflections on the meaning of “risk free”
https://www.bis.org/publ/bppdf/bispap72l.pdf

The idea of risk-free sovereign bonds is best thought of as an oxymoron or as an anomaly of recent history. It is not a useful, necessary or an enduring feature of the financial landscape.

In order to think clearly about “risk-free” we should be specific about the meaning of risk. Risk is deviation from objective and, thus, risk is relative. We each have different objectives and different circumstances so there are different things that can divert us from our objectives. Most financial intermediaries specify a liability, an expected return or an entire benchmark portfolio as their objective and measure risk as deviation from that.

Central banks control the risk-free rate – properly understood as the yield on short-term government bills. This is a fact of financial life...

[Northern Flicker]

Actually, people do talk about the need to have some savings to cover lumpy expenses that a wage income stream won't cover.

Well I agree people should have some liquidity to meet lumpy expenses

I've never seen anyone call a TIPS ladder risky.

Are you sure?

Northern Flicker wrote: ↑Sun Dec 01, 2024 12:51 am
The inflation-adjusted perpetuity is only a risk-free asset if that return schedule matches your liabilities. I could live to 100 with uniform real liabilities up to the time of death. Or I could need 5 years of skilled nursing care starting at age 70 and die at 75. If we define risk as the risk of not bring able to cover our expenses, then the fact that either outcome could materialize shows that the perpetuity is not risk-free

Since a TIPS ladder is the closest thing we can purchase to an inflation indexed perpetuity…

If you define "risky" as anything that is not risk free, then I can see why you came to that conclusion. That is not my definition of risky. In fact, there is no such thing as a risk-free investment as a practical point. TIPS, t-bills, and FDIC or NCUA insurance are still exposed to the risk of a default by the US govt, however remote that risk may be. So if you choose to define risky as anything that is not risk-free, technically speaking, everything would be risky.

Moreover, I actually provided the definition of risk that I care about, which is the probability of running out of assets. That is a portfolio risk measure. A TIPS ladder is a useful tool for managing risk. It can provide a predetermined set of real cash flows with very minimal risk, which can be incorporated into a financial plan to cover that part of the plan at very low risk.

But having a single risk-free investment is not generally even a meaningful goal, because no investment is neatly risk-free in isolation. If you held 100% of your assets in a 30-year TIPS ladder delivering 30 equal real payments, the TIPS ladder likely would become significantly riskier. It only is very low risk when properly integrated into a financial plan and portfolio. It is very low risk at delivering the portion of the plan that it is properly designed for.

[ScubaHogg]

Well I agree people should have some liquidity to meet lumpy expenses



Are you sure?



Since a TIPS ladder is the closest thing we can purchase to an inflation indexed perpetuity…

If you define "risky" as anything that is not risk free, then I can see why you came to that conclusion. That is not my definition of risky. In fact, there is no such thing as a risk-free investment as a practical point. TIPS, t-bills, and FDIC or NCUA insurance are still exposed to the risk of a default by the US govt, however remote that risk may be. So if you choose to define risky as anything that is not risk-free, technically speaking, everything would be risky.

Risk free in the conventional sense of how the word is used in finance when discussing payoffs. So we have something to compare “risky” equities to. It’s the norm.

If we want to redefine words from their conventional usage we can, but it’ll just lead to a lot of confusion without providing any more wisdom. Obviously nuclear war could happen, a meteorite could strike, a civil war could bring down a govt.

But peppering every single conversation with all those caveats is exhausting, pedantic and utterly pointless

[Northern Flicker]

Risk free as used in finance such as when we discuss the "risk-free rate" uses short-term volatility of asset value as a risk measure. A 30-year TIPS ladder is nowhere near risk free in that measure.