Thursday, March 27, 2014
Retirement Income Strategies
The following chart is a first draft of something I hope to develop more. It is a list of different approaches to retirement income which I've come across. I'm still thinking about if there are ways that I could improve the classifications, and I've probably forgotten to include something worthwhile. I'm open to feedback and suggestions.
Very useful. We need an overview of the universe of retirement income solutions.
ReplyDeleteThanks Kerry.
DeleteWade.. thanks for all the info you have supplied. One area that may be a tad different due to WR's ... and I have not seen a portfolio mortality study on, is IRS RMD. To those of us a few years from starting RMD any data on IRA portfolio longevity from AA during RMD would be appreciated.
ReplyDeleteThank you. There are some studies which look at using the RMD rules as a way to guide retirement spending. This is a good starting point:
Deletehttp://crr.bc.edu/working-papers/should-households-base-asset-decumulation-strategies-on-required-minimum-distribution-tables/
Wade, as you dig into variable spending strategies it's worth taking a look at Bob Clyatt's 4%/95% withdrawal system from "Work Less, Live More".
ReplyDeleteIt's been around for nearly a decade (like some other systems) and it's ridiculously simple (unlike many others). It's been examined by some backtesting but I think a rigorous analysis could reassure retirees who find other systems too complicated.
It's also a simple starting point for more ways to model variable retirement spending algorithms.
Thanks Doug. I've heard about that one and forgot. Bob's Financial Website had a great article about different variable withdrawal strategies. His website is gone, but I think I have a print out. I'll aim to get those added.
DeleteHi Wade,
ReplyDeleteThat is a very thorough diagram! I've asked John if he's aware of other work that may fit as well.
Our joint paper coming in April's JFP would be considered a hybrid where it starts out Probability Based and then switches, either fully or partially, to Safety First. Although we used a method in one segment of the chart, there should be little reason other approaches couldn't determine the annual expected cash flows and perform a similar breakeven analysis we developed in that paper as well.
I think that kind of work: when, where, why and how to transition between the two approaches is where valuable work in the future should come from.
A nice summary Wade!
Thanks Larry,
DeleteGood point. I'm not immediately sure how to classify that paper, which will be in the JFP in a few days. Certainly it's variable spending, and maybe a component of at-risk flooring. Hmm...
When considering the big picture of retirement planning strategies, it is necessary to consider the level of uncertainty attached to each of the planning variables, and to what extent each variable is unique to the individual (client). Most strategies contain specific rules and formulas. IMO the most important consideration with respect to any strategy is how it deals with the uncertainties and with the uniqueness of the individual. The acid test is whether the strategy can adapt to these factors, and if it does so adapt, what, if anything, is left of the rules and formulas of the strategy. Do those rules in any way constrict the planning?
ReplyDeleteAnother approach that some may not consider a "strategy" is to concentrate on a detailed analysis of the client's financial position, while incorporating the client's specific personal and psychological attributes. The knowledgeable financial planner should then be able to apply the relevant theory contained in Wade's list of retirement income strategies to their client; considering all market/inflation uncertainties in the context of the unique situation of his client.
It may be that that is as good as it gets.
Thanks John.
DeleteThanks Wade!
ReplyDeleteYou might want to indicate whch have been written about vs those that have been made usable.
ReplyDeleteExcellent, thanks Wade.
ReplyDeleteI wonder whether a "Hybrid Approach" should be a third approach. It seems to be becoming more popular to try and combine some of the best of safety first with the more likely upside of probability based investing. There seem to be several Hybrid categories. Those that lock in guaranteed returns for just the first 10 to 15 years (Dirk Cotton blogged about his approach recently). Those that lock in just a certain percentage eg 50% with TIPS and SPIAs etc.and count on a very safe SWR, eg 2% ,for the rest of the guaranteed "floor". Those (maybe like Larrys) that start with SWR but expect to annuitize by mid 70s or 80.
these categories could be something like
ReplyDeleteTruncated safety
Partial safety
Delayed (or defered) safety
Derek,
DeleteThanks. I think you're doing a good job navigating through some of the key distinctions. The truncated safety is essentially time segmentation. It's a front-end floor that meets the lifestyle spending goal for the short/medium term horizon. Partial safety is essentials vs. discretionary, and this is best represented with the locked-in flooring category. Flooring doesn't have to apply to the entire lifestyle goal.
As for the delayed or deferred safety, this is a good point and is also where I'm having trouble classifying what Larry described. In terms of what is already on the table, the Russell Investment's Funded Ratio Management approach is the closest to this, as it uses probability-based approaches while tracking how funded the client is, and then jumps in to lock in the lifestyle goal safely if an overfunded client's ratio falls down too close to 1. But Larry is describing something which is more like using a variable withdrawal probability-based strategy until the benefits of income annuities become so overwhelming that it's hard not to justify a shift.
Building up on earlier comments:
ReplyDeleteChapter 4 of the RMA Curriculum maps 9 specific household risk exposures into 4 general categories (Political, Business, Behavioral, Chance).
The risk exposure map is then stretched across two dimensions: Frequency of Risk vs. Severity of Risk to produce 4 quadrants that we call the risk management techniques allocations (Retention, Management, Pooling, Avoidance). These risk management techniques allocations are the foundation for the Retirement Allocations: Upside, Floor , Longevity , Reserves.
Safety First vs. Probability First is a distinction developed during an on-line discussion of the RMA Governance and the Curriculum Committees. Credit for inspiring the distinction goes to Harold Evensky. It is a distinction that you extended extraordinarily well, including a comprehensive side-by-side comparison at the end of Chapter 4, on page 295 of the RMA Curriculum book available on Amazon. The distinction helps in the mapping of Implementation Process Approaches (e.g. SWP) with Retirement Allocations portfolio (e.g. Upside).
Implementation Process Approaches are what you label Retirement Income Strategies on your chart.
Chapter 5 of the RMA Curriculum is a curated collection of individual sections that summarize implementation process approaches that vetted authors have documented. Ten of the approaches in your matrix are currently documented in the RMA Curriculum.Should the authors of other approaches be interested in having sections in the RMA Curriculum, do reach out to me.
Please consider the following suggestions for your chart:
- Mike Lonier’s R-Map software is a near complete software implementation of the current edition of the RMA curriculum. Mike developed it to better advise his own clients. R-MAP integrates probability first as well as safety first implementation process approaches using the framework of the Retirement Allocations. We use it to teach the RMA courses.
R-MAP’s position in your chart may be too limiting for its range of capabilities. It uses all 4 Retirement Allocations to create comprehensive client retirement plans, instead of focusing on a specific implementation process approach.
The RMA curriculum identifies three fundamental retirement planning strategies before drilling in the specific implementation process approaches: Investment –based strategies, Goals-based strategies and Product-based strategies. R-MAP implements Goals-based planning.
- Mike Zwecher’s implementation process approach is listed in the curriculum as one of the Engineering Process approaches but it is not known as “The Engineering Process Approach”. We refer to it in the curriculum as “The Zwecher approach”.
It also integrates several columns in your matrix for the same reason's as Mike's. The Zwecher approach implements Investment-based planning.
Wade, your contributions to the RMA were of the highest caliber and your blog continues in the same tradition.
Thank you Francois.
DeleteI'll make updates based on your suggestions.
As you are describing that the Zwecher approach and R-Map should both also extend to the top part of the table, and also there are a few others that bridge the gaps between the top and bottom of the tables (Income Discovery, Tony Webb's work, some of the dynamic programming approaches), there may still be a better way to organize these. Keeping the top and bottom sections separate may not work completely.
Isn't retirement planning really about the judgment and skill of the planner and not about what formulas, rules or strategies he follows? Isn't there but one strategy?
ReplyDeleteHow different investments will perform in the future is unknown, but we should all be able to agree the broad range of possible results. We should be able to classify investments from the safest to most risky and estimate the range of possible results of each -- the safest will have virtually no risk but limited reward, and the riskiest will have substantial risk but much upside. Although future performance of investments is an unknown, we should be able to establish parameters and estimate a reasonable worst case scenario for any mix of investments.
When dealing with a specific individual, we should all be able to determine the broad range of retirement needs. In this context we should be able to determine essential, deemed necessary, discretionary, and (low-risk) catastrophic potential needs. We should be able to estimate a reasonable range of inflation rates and a conservative life expectancy. Again there are many unknowns, but we should be able to establish parameters for a specific individual.
Wade's listed retirement strategies demonstrate the considerable level of accumulated knowledge in this field. The knowledgeable financial planner should understand the theory behind these strategies, and apply this to establishing an investment and spending plan in consultation with the individual. What else is there? Why so many categories and implementation process approaches?
John,
ReplyDeleteGood question.
Looking at the long and growing list of proposed retirement solutions (i.e. the retirement income strategies in Wade’s chart or the implementation process approaches sections in Chapter 5 of the RMA Curriculum) suggests that one size does not fit all clients.
One of the three core features of the RMA Curriculum is the advisor’s ability to create understandable, practical and defensible “process allocations” – The Retirement Allocations – tailored for each household client.
These “process allocations” can be expressed simply in the Retirement Policy Statement using the four Retirement Allocations portfolios: Upside, Floor, Longevity, Reserves.
Advisors familiar with asset allocation (e.g. 60/30/10) find it easy to transition to Retirement Allocations.
For instance, an initial 60/30/10 accumulation portfolio can become a retirement portfolio – typically for a HNW/Over-funded household client - by asset/liability matching the bond portfolio (e.g. Ladders as an implementation process approach/retirement income strategy).
The 60/30/10 asset allocation thus becomes a 60/30/0/10 Retirement Allocation.
This framework can be applied to all of the implementation process approaches/retirement income strategies in the chart and beyond.
The increasing complexity created by the growing number and types of implementation process approaches/retirement income strategies becomes manageable for the advisor and easily communicated to the household client with the Retirement Allocations.
Francois, I can see that there are grounds here for strenuous agreement :).
ReplyDeleteNo disagreement that retirement allocations must be tailored for each household client. In fact, one might conclude that there is an indefinite number of possible allocations given the uncertainty of markets and the uniqueness of each client. As you suggest, excellent advancements have been made in this area, for example, in the area of life-cycle LMP investing.
I believe that the emphasis of each advisor should be on making certain that he fully understands all financial and personal aspects of the client. The advisor then selects the best retirement allocation for that specific client. I suggest that attaching the labels and categories may be constraining and therefore possibly unhelpful. Perhaps a checklist of possible alternatives is helpful for some, but the knowledgeable and skilled advisor may well be better off ignoring all labels and planning from a clean slate for each client.
As I see it, the danger is not seeing the forest for the trees.
This is great! thanks
ReplyDeleteThanks Wade. It serves as a big picture for beginners to learn about the subject. Question: Why does Financial Engines (W. Sharpe) belong to the Safety-First category? From an end user perspective, Financial Engines allowed me to enter my asset allocation, my retirement date, and the length of my retirement. After running simulation on some scenarios, it told me how likely my plan would succeed. It appeared that it took a probability-based approach, at least on the surface. Perhaps I need to look a little deeper under the hood of Financial Engines. I'd appreciate it if you would point me to the right direction to begin understanding the rationale behind your characterization.
ReplyDeleteIssac,
ReplyDeleteThe way you've described it, that does sound like the probability-based approach. I may have misunderstood how Financial Engines works, as I thought it was similar to the Dimensional Managed DC in terms of trying to make a DC pension behave more like a DB pension. In that regard, I thought they guide the asset allocation for you and invest your assets more conservatively until you are sure that you can at least meet your essential needs in retirement. As well, a lot of the research I've seen from Financial Engines employees (Sharpe, Scott, Watson) could definitely be classified as safety-first. I need to do some more digging to see how Financial Engines works though. Thank you.
I would love to see "Lifestyle Driven Investing" listed among your safety first strategies. I have 350 retired clients who have certainly benefitted from this strategy - some for more than 20 years. Hope you are well. I do enjoy keeping up with your work. Erin Botsford
ReplyDeleteThanks Erin. I'll add it in the next iteration.
DeleteWade,
ReplyDeleteThis is very cool, thanks! Over the last couple of years, I have been drinking up information about the safety first approach, and I’m still thirsty for more. I’d love to see something like this organized more like a marketplace (non-commercial) than as a 2-D chart, allowing users to drill down in chosen strategies for more detail. Even better, if users could share their opinions and experience on these strategies as part of the more detailed information, almost like Amazon reviews but of specific strategies rather than products or books. Summaries of pros and cons, star raiding, or “best used when…” would be interesting and potentially useful as various stakeholders’ opinions accumulate for each strategy. I’d be really cool if the minds behind the strategy even participated, not as reviewers but at some level of detail, just like how Amazon has a from the author or manufacturer paragraph in the product description. This may be a bigger project than you had in mind, but nothing like that exists, and it’d be incredibly useful.
Planning to earn after the retirement. You must keep an eye on Forex Signals to earn money even after the retirement.
ReplyDeleteOne of the nice information on Retirement Income Strategies.Thanks for your great information about it.
ReplyDeletehttp://www.retirekit.com/