iTDF - silver bullet for accumulation and decum??

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
tman9999
Posts: 38
Joined: Tue May 20, 2008 1:06 pm

iTDF - silver bullet for accumulation and decum??

Post by tman9999 » Sun Dec 01, 2019 4:45 pm

Anyone investigated this? Can’t find much about it on interwebs, and only one mention of iTDF on B-heads. https://www.ipe.com/pensions/pensions/b ... 83.article
Key points
* iTDFs combine two multi-asset funds, varying the asset mix algorithmically over time
* They can be used in the accumulation and decumulation phases
* The concept aims to provide a smooth income scheme without the provider having to assume mortality risk

User avatar
David Jay
Posts: 7282
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: iTDF - silver bullet for accumulation and decum??

Post by David Jay » Sun Dec 01, 2019 5:48 pm

Are you confident that this is a commercially available product? Reading the article, the last paragraph appears to be a request for financial backing to create these products.
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

retiringwhen
Posts: 1436
Joined: Sat Jul 08, 2017 10:09 am
Location: New Jersey, USA

Re: iTDF - silver bullet for accumulation and decum??

Post by retiringwhen » Sun Dec 01, 2019 6:04 pm

The article is pretty sparse on details, but it sounds suspiciously like a variable annuity embedded in a tax-deferred account. In the US, this would be a usually found in a 403(b) or 401(k). Looks like the usual complex vehicle with that will only complicate valuation of the investments and an opportunity for the seller to create additional fees while selling safety (at a price).

I am basing this on these statements:
iTDFs may also be designed to produce smooth income during retirement, either for a pre-defined period (for example, between ages 65 and 85) or, if combined with optional longevity pooling, for as long as the account owner lives. It is unique to iTDFs that a seamless transition between the two phases are offered.

The retirees may maintain flexibility and control of all of the savings during the liquidity period and they do not have to enter the life contingent phase, for instance because of deteriorating health.

On the other hand, it is expensive to self-insure longevity and there is a compelling case for the benefits of sharing longevity risk between the retirees when they get older. The survivor benefits become more significant the older the retiree gets and the mortality gains for those alive will have more impact than asset returns at older ages.

Atticus
Posts: 20
Joined: Wed Aug 07, 2019 11:43 pm

Re: iTDF - silver bullet for accumulation and decum??

Post by Atticus » Sun Dec 01, 2019 9:37 pm

Agree, it doesn't look like it is commercially available yet. I think my inclination is to suspect that it will be a higher fee, sold-not-bought product. But I'm not sure it has to be; I could see a market for such a product if it's kept understandably simple and low-cost.

Transitioning retirement savings to a steam of retirement income remains a challenge for many who are less DIY-inclined than most BH's. It's also challenging for many to mentally translate a nest egg amount into an equivalent annual income. John Rekenthaler has written a bit about this (https://www.morningstar.com/articles/93 ... ent-income) and I'm sure others have as well.

Hopefully, at some point the major DC plan providers will embrace a target-date equivalent for decumulation. If it has similar features to an SPIA, all the better.

retiringwhen
Posts: 1436
Joined: Sat Jul 08, 2017 10:09 am
Location: New Jersey, USA

Re: iTDF - silver bullet for accumulation and decum??

Post by retiringwhen » Sun Dec 01, 2019 9:49 pm

Atticus wrote:
Sun Dec 01, 2019 9:37 pm
Agree, it doesn't look like it is commercially available yet. I think my inclination is to suspect that it will be a higher fee, sold-not-bought product. But I'm not sure it has to be; I could see a market for such a product if it's kept understandably simple and low-cost.

Transitioning retirement savings to a steam of retirement income remains a challenge for many who are less DIY-inclined than most BH's. It's also challenging for many to mentally translate a nest egg amount into an equivalent annual income. John Rekenthaler has written a bit about this (https://www.morningstar.com/articles/93 ... ent-income) and I'm sure others have as well.

Hopefully, at some point the major DC plan providers will embrace a target-date equivalent for decumulation. If it has similar features to an SPIA, all the better.
Vanguard's Managed Payout fund was an attempt and went over like a wet noodle.

I guess a LifeStrategy Fund could have a version that is set to a VPW like withdrawal and do monthly distributions of dividends and capital over time (with an assumed duration) to simulate such an approach. You would need to create new ones every year or so I guess to make them evolve. I still can't see them being anymore more popular than the payout fund.

Topic Author
tman9999
Posts: 38
Joined: Tue May 20, 2008 1:06 pm

Re: iTDF - silver bullet for accumulation and decum??

Post by tman9999 » Sun Dec 01, 2019 10:07 pm

I thought this summed it up nicely, and is what I’ve been anticipating will become the norm. Not there yet, but I think this guy’s on the right track.
Relying on a robust algorithmic framework, iTDFs will fit easily into an increasingly digitalised and mass-customised world and can be delivered as fully automated solutions. The algorithm-based product design allows scalability, portability and low cost.
1. I think the notion of an algorithimically-driven ETF (if in fact that’s what this is proposing) is a realistic next iteration of FinTech and the application of robos.
2. He’s proposing a two-fund portfolio, in effect, built around a model that includes flexibility to accommodate different life stages and risk tolerance levels. Again, seems very doable from a technology perspective.
3. If that is doable, how is that much different than the BH 3-fund portfolio approach? One person commented that this might be of interest to investors who lack the level of sophistication of many of the BH forum members. Does that mean that most BH forum members don’t use the 3-fund portfolio approach? Or does it mean that a 3-fund portfolio approach does represent a level of sophistication that is beyond the average investor? If the latter, then why wouldn’t the two-fund iTDF portfolio approach represent a sophisticated approach? And if the former, then I’m curious as to why BH people are advocating the 3-fund or 4-fund approach, but not actually following it themselvess.

Topic Author
tman9999
Posts: 38
Joined: Tue May 20, 2008 1:06 pm

Re: iTDF - silver bullet for accumulation and decum??

Post by tman9999 » Mon Dec 02, 2019 10:49 am

David Jay wrote:
Sun Dec 01, 2019 5:48 pm
Are you confident that this is a commercially available product? Reading the article, the last paragraph appears to be a request for financial backing to create these products.
Agree. This is written by a Danish university professor who is working on making this available in Denmark, and advocating it conceptually as a solution option in general. I think he’s on the right path. Was curious what others here thought, and whether anyone was aware of similar offerings available in the US (eg, as someone mentioned above, Vanguard’s offering, which apparently didn’t go over that well).

User avatar
David Jay
Posts: 7282
Joined: Mon Mar 30, 2015 5:54 am
Location: Michigan

Re: iTDF - silver bullet for accumulation and decum??

Post by David Jay » Mon Dec 02, 2019 11:13 am

tman9999 wrote:
Mon Dec 02, 2019 10:49 am
Was curious what others here thought...
The devil is always in the details. There is no product and thus no details.

It’s hard to have an informed opinion without information.
Prediction is very difficult, especially about the future - Niels Bohr | To get the "risk premium", you really do have to take the risk - nisiprius

retiringwhen
Posts: 1436
Joined: Sat Jul 08, 2017 10:09 am
Location: New Jersey, USA

Re: iTDF - silver bullet for accumulation and decum??

Post by retiringwhen » Mon Dec 02, 2019 11:32 am

tman9999 wrote:
Sun Dec 01, 2019 10:07 pm
I thought this summed it up nicely, and is what I’ve been anticipating will become the norm. Not there yet, but I think this guy’s on the right track.
Relying on a robust algorithmic framework, iTDFs will fit easily into an increasingly digitalised and mass-customised world and can be delivered as fully automated solutions. The algorithm-based product design allows scalability, portability and low cost.
1. I think the notion of an algorithimically-driven ETF (if in fact that’s what this is proposing) is a realistic next iteration of FinTech and the application of robos.
2. He’s proposing a two-fund portfolio, in effect, built around a model that includes flexibility to accommodate different life stages and risk tolerance levels. Again, seems very doable from a technology perspective.
3. If that is doable, how is that much different than the BH 3-fund portfolio approach? One person commented that this might be of interest to investors who lack the level of sophistication of many of the BH forum members. Does that mean that most BH forum members don’t use the 3-fund portfolio approach? Or does it mean that a 3-fund portfolio approach does represent a level of sophistication that is beyond the average investor? If the latter, then why wouldn’t the two-fund iTDF portfolio approach represent a sophisticated approach? And if the former, then I’m curious as to why BH people are advocating the 3-fund or 4-fund approach, but not actually following it themselvess.
I think that tools like Vanguard's PAS come very close to providing the solution you are describing. Once someone starts to think about the massive set of parameters that actually go into a simple retirement portfolio withdrawal strategy, you realize that the number of ETFS required is nearly infinite. Just consider the items I can think of in one minute.

1.) Social Security / Pensions available to the investor
2.) Joint vs. Single situations
3.) Expected Longevity (age, health, etc.)
4.) Risk Tolerance
5.) Legacy Motive
6.) non-investment assets that may impact income needs such as disability or veterans benefits.
7. ) Real Estate holdings.
8.) Mix of different tax-advantaged account configurations
9.) Funds that the investor has little or no control over (e.g., limited 401(K), trusts, etc.)

It seems like someone would need to create an algorithm to munge those variables and turn it into a key that maps to one of the Target Date like iDTF envisioned by the author. I can see the need for hundreds of ETFs just to give a reasonably fine-grain mapping from those requirements. By comparison, there are only about 11 Target Date funds needed for most accumulation setups.

Besides, none of those variables above stay constant over the long-term and that means choosing to make changes along the way.

A tool like PAS, or the various approaches from Betterment, Schwab, etc. all take the same variables and then map them into a key and produce a model portfolio.

Of course, at 15bp, PAS is still not free, but it is getting pretty darn cheap. the non-human version is now down to 10bp! I can't imagine these iTDFs being run for any less, especially if they have to create hundreds of them!

Topic Author
tman9999
Posts: 38
Joined: Tue May 20, 2008 1:06 pm

Re: iTDF - silver bullet for accumulation and decum??

Post by tman9999 » Mon Dec 02, 2019 12:16 pm

retiringwhen wrote:
Mon Dec 02, 2019 11:32 am
I think that tools like Vanguard's PAS come very close to providing the solution you are describing. Once someone starts to think about the massive set of parameters that actually go into a simple retirement portfolio withdrawal strategy, you realize that the number of ETFS required is nearly infinite. Just consider the items I can think of in one minute.

1.) Social Security / Pensions available to the investor
2.) Joint vs. Single situations
3.) Expected Longevity (age, health, etc.)
4.) Risk Tolerance
5.) Legacy Motive
6.) non-investment assets that may impact income needs such as disability or veterans benefits.
7. ) Real Estate holdings.
8.) Mix of different tax-advantaged account configurations
9.) Funds that the investor has little or no control over (e.g., limited 401(K), trusts, etc.)

It seems like someone would need to create an algorithm to munge those variables and turn it into a key that maps to one of the Target Date like iDTF envisioned by the author. I can see the need for hundreds of ETFs just to give a reasonably fine-grain mapping from those requirements. By comparison, there are only about 11 Target Date funds needed for most accumulation setups.

Besides, none of those variables above stay constant over the long-term and that means choosing to make changes along the way.

A tool like PAS, or the various approaches from Betterment, Schwab, etc. all take the same variables and then map them into a key and produce a model portfolio.

Of course, at 15bp, PAS is still not free, but it is getting pretty darn cheap. the non-human version is now down to 10bp! I can't imagine these iTDFs being run for any less, especially if they have to create hundreds of them!
+1 on the complexity of decumulation modeling. I’ve been wrestling it for the last couple of years, and created a list of decisions that includes everything on the one you just brainstormed, plus a few more.

Timing is a big variable I’d add to your list. When to take SS for each person in a married couple is nontrivial. I’ve created my own model for helping me decide (I know - lots of people have already done that).

Then some crystal ball gazing to determine whether/when/how much tIRA to ROTH we should convert, based on anticipated tax rates - and whether current rates are going to go up (probably), and by how much (who knows?). Etc etc etc.

Which is why I’m confident that it’s only a matter of time before someone comes with a cost-effective AI/ML-based approach to the decumulation problem. At 15bp I’d consider it, especially if it included predictive modeling of my glide path, with SWR updates accordingly.

BTW, I looked at Vanguard’s PAS site and it says their rates are 30bp for up to $5m, and then 20bp for $5-10m. That’s less than some robos, but still too much money imho. I think it’s entirely possible (probable?) that this problem is going to be solved using smart algorithms and priced in a way that’s not much different than any other SaaS-based service: very competitively, and less expensive than any existing robo out there. Which is why I’m interested in this iTDF in the first place.

Post Reply