Wednesday, June 18, 2014

Kitces and Guyton on the Three Approaches for Retirement Income Planning

Jonathan Guyton and Michael Kitces are both leading advocates of systematic withdrawal strategies for retirement, which involve investing a portfolio using a total returns perspective and having a withdrawal strategy in place to spend down the portfolio sustainably over retirement. This approach is also known as 'safe withdrawal rates.'

In the following video from the New York Life Center for Retirement Income at The American College, they discuss systematic withdrawals, as well as time segmentation and essentials-vs.-discretionary.

As academics tend to favor approaches related more closely to essentials-vs.-discretionary, it is somewhat uncommon to hear rigorous intellectual defenses of systematic withdrawals. Guyton and Kitces do an admirable job in this video, making it relevant for individuals on both sides of the debate. They defend systematic withdrawals, and they focus particularly on explaining why they think time segmentation and essentials-vs.-discretionary are not better solutions.

21 comments:

  1. Excellent video Wade ... kudos to Michael and Jonathan for their explanations!

    ReplyDelete
  2. This conversation doesn't address important questions: How well funded is my retirement? Do I need the extra juice from mortality credits or can I live without them? Can I sleep at night without an adequate cushion of guaranteed income? Do I value legacy over income? SWiP is for the 5% or 10% who can afford it or tolerate the risk of it. As for the last argument, that annuitization means a kind of settling for a safer but lower level of income (it was the argument used against indexing), I think it's not true unless people over-annuitize. Can you overannuitize? Of course. But that's a straw man. But it's not enough to ask if a strategy is right or best. You have to ask, for whom? And under what circumstances?

    ReplyDelete
    Replies
    1. Kerry, thanks for the feedback. All good questions and comments. About that issue of locking a lower lifestyle with an income annuity. I think it's a matter of whether one's lifestyle spending goal requires spending more or less than the annuity payout rate. If they want to spend less, then the annuity can actually help to even support a higher potential legacy value at death. But if they want to spend more, then we are into the issue that their entire spending plan is risky, but the strategy that would give the highest probability to meet the overall spending plan (but which would also create more downside risk) is to maintain an aggressive portfolio of financial assets, though perhaps with some deferred income annuity.

      It always seems to boil down to upside vs. downside, which different people evaluate differently. I do agree with you that lots of Americans will not find the safe withdrawal rate strategy to be comforting and will likely enjoy retirement more with some income annuities.

      Delete
    2. I think Kerry nails one of the key issues in those last few statements. If a client had $100,000 or less saved for retirement, I wouldn't recommend investing in stocks or using SW. At the other extreme, two of my colleagues from the dotcom years are billionaires. Systematic withdrawals would probably work just fine for them. The question isn't "which is better", but "which is better for you?"

      Delete
  3. At 14:26, Guyton talks about a study involving over 400 planners regarding what proportions of their clients had to significantly alter their lifestyle because of the financial crisis. Do you have a reference for that study? The speakers discuss the fact that a higher proportion of clients using a essentials/discretionary policy reduced their lifestyle than of those using the other strategies. But it seems to me that there are several possible explanations for that other than a failure of the strategy. For example, it seems plausible that a higher proportion of those with lower accumulations would be attracted to that strategy because they are in the most danger. If so, perhaps many of them would have needed to retrench even if they had used other strategies. Alternatively, perhaps those using other strategies should have retrenched but didn’t and the bad effects won’t be apparent until some time in the future. Perhaps I would agree with the speakers if I had more information, but it’s just not clear to me what these results mean.

    ReplyDelete
    Replies
    1. It was a 2011 survey of financial advisors conducted by the Financial Planning Association.

      I can't find any public link to it, but I do have a PDF copy. You can obtain it if you have membership at the FPA and access the archives of the Journal of Financial Planning.

      You are making good points, but I don't think the survey will provide you the necessary information to test these ideas. I think the questions were just in terms of asking which strategies advisors used, and then how many of their clients had to make significant cuts to spending after 2008.

      Delete
  4. I think this was an excellent discussion. I have not yet retired (coming soon!), but I am definitely one of those people who is more comfortable with the "buckets" approach. I think there are a few reasons:

    1) Most published data will say "we've done studies that show you can spend X% +/- Y% for Z years (so maybe 5% adjusted for inflation, but you have to lower total spending 2.5% on "bad" market years, and your money has an 80% probability of lasting 30 years." But that's not very helpful. What I need to do is spend 8%, adjusted for inflation, for 10 years, then social security kicks in. Then I want to spend (say) 3% from my portfolio, adjusted for inflation for 10 years. Then I want to spend that same 3%, but NOT adjusted for inflation for the rest of my life. Most studies (and tools) don't support that kind of analysis. But breaking your portfolio into buckets does.

    2) I am fairly comfortable that over the long run, equities will provide the best returns. But I certainly worry about that first few years after retirement where a big market correction really devastates your portfolio. By putting your fixed income into the first "bucket" and your equities into the next (one or two), you buy yourself time where the equities have a chance to get their "time advantage." It also tends to give you the U-shaped asset allocation vs time - heavy equities during work years, convert to lighter equities to fund your first bucket, then as you spend down the fixed income in your first bucket, the equity percentage rises again.

    3) Finally, regarding their study that showed that people with the bucket approach were the ones that found their lifestyle changed the most in the 2008 crash - I suspect (but don't know, of course) that this is because those people had already allocated those funds for non-essentials, and were prepared to cut back when the market went south. Those people that didn't allocate their funds in that manner may have just thought "hey, I had an 80% chance of success starting this thing out, I'm going to stick to my plan and keep spending." That doesn't mean they SHOULDN'T have cut back, they may just have increased their risk.

    ReplyDelete
    Replies
    1. Thanks. Just a few points based on your numbered comments:

      1) you could still do this type of variable spending with a systematic withdrawal strategy. Your spending can vary over time, but the important aspect would be to invest from a total returns perspective rather than as buckets. Nonetheless, I agree in can be easier to think about things with these buckets. Total returns investing can be abstract.

      2) Moshe Milevsky has put together the best example I've seen comparing bucketing to total returns, and it turns out to be random about which strategy actually does better. It depends on the specific sequence of returns in retirement. So this means either strategy can be made to work, and it's just a matter of which you feel most comfortable with.

      3) You may well be right. I don't think the survey dug deep enough to address this.

      Delete
  5. Isn't the bucket approach a rising equity glide path approach by a different name?

    ReplyDelete
    Replies
    1. It could be... that's where Guyton's point about "is there any there there" comes into play. A lot of bucketing strategies seem to not be clear about what to do in the future. If you let the bond ladder shorten as time moves forward, you could end up with a rising equity glidepath. But if you always extend the bucket out, it may not be a rising glidepath, depending on how equities perform relative to the costs of extending the bucket.

      Delete
  6. Have Jonathan and Michael debated the merits of bucket approach with Harold Evensky, who is a champion of the bucket approach ? Evensky said that he has implemented the bucket strategy in his firm for over at least of couple of decades and he said it has worked very well with his clients

    ReplyDelete
    Replies
    1. I've too referred to the Evensky and Katz cash reserve strategy as a bucketing approach. But Harold disagrees with this classification. He prefers to think of it as total returns investing with a risk management overlay for having a few years of front-end expenses held in cash. I'm not aware of either Jonathan or Michael's thoughts about this, but I can ask them. This seems like something Michael may already have a blog post about. I'll send Jonathan an email about it.

      You might want to see this column:

      http://www.advisorperspectives.com/newsletters12/Retirement_Floors_and_Implications_for_Evenskys_Cash-Reserve_Strategy.php

      Actually, in the first draft I referred to the cash reserve strategy as a bucketing approach, and that is when Harold and I talked, and he explained how he thought about it differently.

      Delete
  7. A great discussion. They make an excellent case for simple over formulaic systems; now if someone would make the best case -- eliminate all systemic approaches.

    Financial advisors must understand three things: 1) they must understand the fundamentals of retirement investing as explained in your 6/20 book review; 2) they must understand the range of possible investment outcomes during retirement by understanding economic fundamentals , and to a lesser extent from history; and 3) they must fully understand the needs, goals and psychology of each specific client. Each of these three things are essential, but there is no system -- SWR, time segmentation, or essentials-vs-discretionary -- that can be considered "best science".

    Of course some elements of the various techniques discussed in the video might have value in a specific client situation, but IMO it is wrong to claim that any system is best and/or better than any other. The financial advisor must apply his understanding of the three things listed above to each specific client situation and use his best judgment in developing a plan. The plan must be continually reviewed and updated as necessary. Although some advisors may prefer a specific system as a crutch or rationalization, that is the wrong way to proceed, IMO.

    ReplyDelete
  8. That was a good presentation and discussion, and I just wanted to make a couple of points. There was quick reference to a slide that showed immediate annuities (SPIAs) with various bond options (like ladders) and the remark that they all do pretty much the same thing. Clearly they don't because SPIAs provide the benefit of mortality pooling that the bond options do not provide. Based on some recent research I've done, SPIAs do this without the heavy expense loads that impact other annuity products.

    I'm still a fan of floor and upside and using SPIAs to build a higher floor than optimized SS alone provides. People will likely own some bonds anyway, so I like the idea of SPIAs as bonds with the additional benefit of mortality pooling.

    Re: the FPA survey cited--that survey was done in September 2011 and the question asked was about clients who "needed to significantly alter their lifestyle in the past 12 months due to economic conditions." The "past 12 months" in this case had stocks returning over 16% and bonds returning over 4%, so I don't see how that survey is relevant to the systematic withdrawals versus floor-and-upside discussion.

    ReplyDelete
    Replies
    1. Joe,

      I'm trying to look into this more. I found a copy of the May 2011 report of a survey conducted in September 2010. This report indicates the the survey has been annual, and there must have been one conducted in September 2009. That must be the relevant one.

      I've now also found the questionaire for the 2010 survey. None of the questions are about whether clients had to reduce spending.

      I can't find the 2009 survey in the FPA archives.

      I'm stumped on this matter.

      Delete
    2. Wade,

      I've got the 2010 survey and question 7 is the one about clients adjusting lifestyle in the past twelve months. It's a similar question in 2011 except split into pre-retirees and retirees. I also have the 2008 survey (but not 2009) and 2008 doesn't have the question. The write-up of the 2010 summary by Carly Schulaka has an extensive quote from Jonathan Guyton and he wrote the whole article for 2011 himself. In both the 2010 quote and much of his 2011 article he argues for systematic withdrawals as less disrupted by "an external event (the economic and market conditions of the moment)" than essential/discretionary. But the 2010 and 2011 surveys both refer to the "past 12 months" in both cases the markets did well, so I can't understand how the surveys support his argument.

      It's also worth noting that systematic withdrawals as defined for the FPA survey includes living off dividends and interest and not dipping into principal. I suspect, but cannot prove, that those using the discretionary/essential approach are less wealthy and more easily disrupted than those using systematic withdrawals as defined for the FPA survey. I think we have to be careful about drawing comparisons without controlling for wealth levels.

      Delete
    3. I did some digging and found some figures from the 2011 survey on wealth levels and choice of retirement income strategies. Among planners who frequently or always used systematic withdrawals the peak percentage use was by planners with clients having a typical retired assets range of $1 million to $2.49 million. Among planners who frequently or always used essential/discretionary the peak percentage use was by planners with clients having a typical retires assets range of $50K to $149K.

      Delete
    4. Joe,

      Thanks for the continued insights about the survey. This could be a good Advisor Perspectives column for you.

      Delete
    5. Good idea, Wade. I think there's definitely a potential column emanating from this. It's interesting to sort out the differences in financial impact from various strategies versus which strategy may work best for a particular advisor in terms of dealing with clients--much more subjective. Sometimes these two aspects get confused.

      Delete
    6. That data about how choice of WD system may relate to portfolio size is great, and I certainly second Wade's suggestion about the worth of an article on it.

      Delete
  9. The observation that you might end up with essentially the same portfolio whether you use SW or buckets is mildly interesting, but misses the point. We could also treat household budgets as having just two categories: "money coming in" and "stuff I have to pay for". We break these into subcategories (food, housing, etc.) for the same reason we need buckets, to make our finances more transparent and more manageable. Though, I agree that ultimately we have the same amount of money coming in and stuff to pay for.

    Also, that next bucket doesn't just "deplete at an inopportune time". It rolls in one year at a time. Every year we readjust to conditions as best we can. We roll over a year of bond ladder. We shorten the number of years the most distant bucket needs to finance. I don't see the issue.

    Lastly, being disappointed by not having the retirement I want is better than being disappointed by becoming destitute. But every retiree gets to make that choice.

    ReplyDelete