Asset Allocation Mo...
 
Notifications
Clear all

Asset Allocation Mode-3

20 Posts
6 Users
0 Likes
333 Views
(@hecht790)
Trusted Member Customer
Joined: 4 years ago
Posts: 52
Topic starter  

Most advisors recommend using Asset Allocation (AA) similar to PRC Mode-2. i.e., defining the ratio between equity and fixed investments. 60/40 (equity/fixed) is a popular ratio and usually the recommendation is to increase the fixed part and decrease the equity part as we age (50/50 and then 40/60, 30/70 and even more). The idea is that as we age, we should be more conservative and take less risk since we have less years to recover from a long stock market decline. The equity part of the AA ratio is the risk/reward, and the fixed part is the risk protection. My allocation is 33% US equity, 33% International equity and 33% US bonds. So, my Asset Allocation is 66/33.

If the overall portfolio is big enough so the Fixed part can protect the investor for many years, maybe 10 or more without the need to touch the equity part, then the investor may consider the following alternative:

Asset Allocation Mode-3 proposal:

Instead of Mode-2 static ratio (equity%/Fixed%), use constant $ amount (+inflation) for the Fixed part year-to-year and let the equity part be the rest without % limitation. The fixed $ amount should be big enough to cover many years (10+?) of expenses. For people with a portfolio that grows more than their expenses the effect is the opposite of Mode-2. The older you get the larger the % of your equity be, and smaller the % of your Fixed.

I am not proposing to replace Mode-2. I think Mode-2 is the basic and most people should use it especially if there is a concern of running out of money. But Mode-3 is useful for people with large portfolio that want extra return with relative minimum extra risk.


   
ReplyQuote
(@chrisb)
Active Member
Joined: 4 years ago
Posts: 8
 

@hecht790 Interesting idea. Would defining time periods with different allocations help in this direction? I use Build > Financial Assets > Advanced Portfolio Modeling, Tab: Portfolio Time Periods (and, all the associated tabs specifying what's *in* those periods) to manually accomplish something akin to what you're describing (I think).


   
ReplyQuote
(@hecht790)
Trusted Member Customer
Joined: 4 years ago
Posts: 52
Topic starter  

@chrisb

It is difficult to implement Mode-3 by using the current Mode-2. You will need to change the equity/fixed ratio in each time period as your portfolio grows (or shrinks) to keep the Fixed $ amount constant. It is a kind of iterative process since each time you change the equity/fixed ratio the portfolio size will change, and you will need to correct the ratio again. A built-in Mode-3 will not need to use time periods, just a single definition of the Fixed $ amount. A hybrid of mode-2 and Mode-3 is also possible. It will have time periods, and the user will have the option to use equity/fixed ratio or specific Fixed $ amount in each period.


   
ReplyQuote
(@ricke)
Estimable Member Customer
Joined: 4 years ago
Posts: 174
 

I'm not sure you would really want to follow that strategy. If you maintain the # of years in the fixed income bucket, then if we get a 2022 when both stocks and bonds are down, you would have to sell stocks while they are down to buy bonds, so your % bonds would be going up and % stocks would be going down while stocks are down. I don't think I've ever heard anyone recommend doing that.

Or are you really looking for a bucket strategy that spends bonds when stocks are down? Those are not generally going to work out better than a fixed asset allocation (other than by luck) and buckets require multiple rules about emptying and refilling, with lots of possible rule choices that could be made.

Another quirk happens if you are running out of money, such as if you are doing Consumption Smoothing (where you intentionally trying to spend down to zero over your lifetime). You would be eliminating stocks rapidly as you spend your assets down and would be 100% bonds several years before the end of the plan. That's not generally recommended either.

My wish list would be to have the program do a glide path when you enter new allocations instead of making giant steps. For instance, near-retirees build up their bonds a little at a time, usually as they have money available, so it would be much more natural to have the program do that rather than make one or two big dumps of stocks to buy bonds that might result in the program finding phantom capital gains taxes.

Another instance where a glide path would help is for correctly calculating the optimum SS claim date. Right now, the SS Claim Date Optimizer treats your assumed growth rate in stocks as being no more risky than inflation-adjusted guaranteed income from SS. But SS is much closer in risk to bonds (actually long TIPS) than to stocks, so what folks should do is follow the best risk-adjusted plan and spend down bonds while waiting to claim SS. Having a glide path could approximate that and would then give reliable answers from the Optimizer. Currently the Optimizer favors claiming too early because it doesn't allow you to preferentially spend bonds while waiting.


   
ReplyQuote
(@hecht790)
Trusted Member Customer
Joined: 4 years ago
Posts: 52
Topic starter  

@ricke

You are right, there are big holes in the proposal. So, Mode-3 as I suggested is not good.


   
ReplyQuote
(@jkandell)
Trusted Member
Joined: 3 years ago
Posts: 61
 

@hecht790 I do something similar. I make sure I keep the NPV of all future essential expenses in a conservative AA (10% stocks), and invest the rest at a fixed AA (in my case 50/50). I think this is pretty similar to what you do, except you only keep 10y worth of expenses, and go 100% on "the rest".

For better and worse, Pralana isn't suited to this type of variable AA, though the glide path can be estimated.


   
ReplyQuote
(@pdxcess)
Active Member
Joined: 11 months ago
Posts: 10
 

I would like to endorse Richard Eaton’s wish list request for a way to allow for a rising equity glide path by allowing for more time periods in which one could change the desired asset allocation.

A related request is one I suggested back in December 2024, but, being a newbie, I mistakenly put it in the Excel section of the forum, instead of here. That request is to be able to specify one TIRA as the ONLY one from which Roth conversions would be made, and when that account was empty, no more conversions. The other TIRA would be the one from which RMDs would start to be taken (though of course the RMD amount would be calculated based on the balances of both TIRAs). When that account was empty, RMDs would then be taken from the “only TIRA from which Roth conversions can be made” account, assuming that account still had a non-zero balance.

Having this ability to specify accounts this way would solve the most limiting aspect of the Roth optimization, which will recommend higher-than-optimal conversion recommendations due to a typically higher stock allocation in the Roth than in the TIRA, and, correspondingly, a higher ROR. With the ability to specify accounts in this way, you could put 100% equities in the only TIRA from which conversions could be made, and fixed income in the “take RMDs from this one first.” That way the ROR in the account from which conversions were coming could match the ROR of Roth, assuming Roth was set to prioritize stocks.

More allowable time periods would allow equity % to rise as fixed income in the TIRA came out as RMDS and was spent.


   
ReplyQuote
(@ricke)
Estimable Member Customer
Joined: 4 years ago
Posts: 174
 

@pdxcess

I solve the different asset allocations in IRA vs taxable/Roth by using Mode 2, have you tried that? That specifically holds the portfolio level asset allocation constant, which seems like the heart of your issue?

My additional desire, regardless of Mode 1 or Mode 2 is to glide rather than just steps. If you think about what the program has to do, it already has to look up the dates of your requested changes and compare to each year's date to select the right allocation. It seems simple to have the program also look ahead to the next asset allocation change and pull that value and date in and interpolate between the anchor point dates and values to get each year's target allocation - of course since I'm not the one that has to do the work, it will seem simple to me!


   
ReplyQuote
(@pdxcess)
Active Member
Joined: 11 months ago
Posts: 10
 

@ricke Yes, I use AA Mode 2, but here’s why I think it doesn’t solve the problem I am describing.

In mode 2, for a given time period, I have to specify 3 parameters for the period: (1) overall portfolio AA; (2) target AA for each account; and (3) account prioritization, which, as I understand it, tells Pralana the order in which to adhere to the target AA for each account, consistent with achieving the overall portfolio AA. (This third parameter is invariant across time periods, but that has nothing to do with the issue at hand.). In my simple case, I have 3 accounts, prioritized as follows: Roth, Taxable, Tax Deferred. Pralana can match my target allocations for the Roth and Taxable for all time periods, but in some time periods, it has to deviate from the target AA for tax deferred in order to maintain the specified overall portfolio AA. So far, so good.

The problem comes when using the optimize Roth conversions feature. I think the problem arises because the the ROR for Roth is higher than the ROR for tax deferred, which is because the ROR is weighted by asset class, and my Roth is 100% stock, whereas my tax deferred has a substantial % of bonds. Pralana sees the higher ROR in Roth and wants to pour conversion $$$ into Roth, regardless of the tax consequences.

My wish list request would solve this problem by allowing the user to set up 2 different tax deferred accounts - one for stocks, and one for fixed income - and tell Pralana: Ignore the fixed income tax deferred account when optimizing Roth conversions. That way, conversions would be done on “from stock only to stock only” basis, and the optimization would not be skewed by differential RoR effects.


   
ReplyQuote
(@ricke)
Estimable Member Customer
Joined: 4 years ago
Posts: 174
 

@pdxcess

Mode 2 is built specifically to solve the problem you are having with different holdings (and therefore different rates of returns) in different accounts, so I'm confused as to where you are having a problem.

I simply set the overall asset allocation and then set the asset allocation of taxable and Roth to be 100% stocks and tax deferred to be 100% bonds, set the order to stocks in taxable 1st, then Roth, then tax deferred. The program works out the asset allocation in the various accounts over time to maintain the portfolio asset allocation, which is what you want to do for Roth Conversion studies.

Note that I put taxable as the highest priority for holding stocks because Pralana doesn't track the tax costs that might be associated with annual rebalancing in taxable.


   
ReplyQuote
(@pdxcess)
Active Member
Joined: 11 months ago
Posts: 10
 

@ricke, Yes, I am using exactly the same asset allocation and account prioritization parameters in mode 2 as you, but here’s the problem I’m having with what I see. When I optimize Roth conversions, I get a conversion amount in year 1 of retirement (2025) of nearly $600k, putting me at the top of the 35% bracket. Without conversions, I would be near the top of the 22% bracket. In addition, the combined federal and state marginal rate of 45% in the “optimized” conversion is higher than the effective/terminal rate of 32% I have set for the rate at which my heirs will be taxed on the balance of my TIRA when I die. When, instead of optimizing, I cap Roth conversions to the top of the 24% bracket (assumed to be 28% starting in 2026), my RMDs are low enough to keep me below the 35% bracket.

So it seems to me that Pralana is telling me to make this massive conversion not because of tax rate arbitrage, but because it is chasing the higher return in the Roth that is 100% stock, which I get is what is designed to do.

My wish list request is intended to enable the user to optimize based on the tax rate arbitrage benefits of conversion only, without skewing the results in favor of conversions due to the higher ROR in the Roth.


   
ReplyQuote
(@pdxcess)
Active Member
Joined: 11 months ago
Posts: 10
 

@ricke I should also add that I am proposing a retirement account whose treatment would be conceptually very similar to the treatment already afforded to HSAs and 529s. Those accounts are exempt from asset allocation in mode 2 - that is, they get whatever allocation you specify for them, and that allocation does not change and exists without regard to the overall specified asset allocation.
I would like a TIRA that is treated the same way and that is blocked from being able to be converted to a Roth.


   
ReplyQuote
(@ricke)
Estimable Member Customer
Joined: 4 years ago
Posts: 174
 

@pdxcess

If the program is really snooping for the higher return of holding stocks, you would see it in the overall asset allocation. If that is OK, then it's not an asset allocation issue, it's more of a question of whether that is truly the optimum and if so, by how much.

Conversions give you a benefit of reducing tax drag in taxable starting now. Later, it not only reduces taxes on RMDs, it decreases tax drag as those RMDs start piling up in taxable. So conversions will win even if you pay a few more percent taxes now than you would later, as long as "later" is long enough later and you didn't pay a fortune in capital gains to pay Roth Conversion taxes.

You could also check to see if you are telling the program that TCJA will expire at the end of the year; if that is your belief, you are giving a boost to more conversions this year.

If you are married, but told the program that either you or your spouse will have a lot of years of paying taxes as a single, that can drive the calculation too.

It's also possible that it is an "optimum", but not one that you should choose. For instance, maybe there's very little benefit for the bulk of the conversion or maybe it is seeing the conversion pull you just below an IRMAA tier for several years, when that is actually too hard to predict very far in the future. You can test for this by noting the final effective estate value, manually making a change and noting the new final effective estate value.


   
ReplyQuote
(@pdxcess)
Active Member
Joined: 11 months ago
Posts: 10
 

@ricke I understand and appreciate all your points - thank you. I nevertheless continue to believe that it would be worthwhile for Pralana to add the feature I have requested: the ability to specify a TIRA with essentially the same parameters that HSA's and 529's already have: Assets in the account do not count toward asset allocation; account is not available for conversions to Roth. This feature would be highly beneficial to users who would like a rising equity glidepath that is not due to reallocating between stocks and bonds but rather that is due to spending down a bond ladder.


   
ReplyQuote
(@ricke)
Estimable Member Customer
Joined: 4 years ago
Posts: 174
 

@pdxcess

Charlie is working on a way to enter a bond ladder or TIPS ladder, where it is assumed that the bonds come available at maturity and do not fluctuate in value other than the planned amount when purchased. My understanding is the focus is on doing this in taxable.

It's an interesting point that many people hold their bond ladders in tax deferred. The question then is what would have to be added/changed to incorporate that. For instance, the program would need rules on what to do when the ladder is the remaining source of funding or the remaining source of RMD money due to poor market returns, high spending or a withdrawal strategy that requires spending assets down. Also, as you suggest, some may want it walled off from doing Roth Conversions. Beginning to sound complicated to me, but stay engaged here as Charlie and Stuart want to solicit feedback on implementing bond ladders to meet customer needs.


   
ReplyQuote
Page 1 / 2
Share: