30

It seems to be common knowledge that a younger person should invest their retirement funds in growth stocks since they can ride out any market swings. It is also common knowledge that as a person approaches retirement age they should tone it down a bit, investing more in bonds, for example.

Why is this true?

I just ran an online actuarial life expectancy calculator and a person 63 years old today can expect, on average, to live to 87. I put in my pretty average "good health" criteria and I jumped to 93 years.

That's 25 to 30 years. It seems to me that, especially in the early decade of retirement, that you should keep a "risky" portfolio, to maximize your retirement income.

DrHouseofSQL
  • 111
  • 4
Paul Cezanne
  • 731
  • 1
  • 6
  • 13
  • 3
    From everything I have read, life expectancy in the U.S is closer to 79 (76 for men and 81 for women). A life expectancy of 87 is rather optimistic, to say the least. – Bob Baerker Mar 27 '18 at 22:10
  • 50
    @BobBaerker That 79 number includes deaths prior to 63. Someone already 63 years old, with no medical issues, would have a longer life expectancy than the average newborn, given the chance of dying in the intervening years. – Grade 'Eh' Bacon Mar 27 '18 at 22:34
  • 3
    @BobBaerker Exactly what Grade 'Eh' Bacon said. Life expectancy varies by age. The numbers you are thinking about are probably from birth, which are the headline numbers often used to describe life expectancy in general. But knowing somebody's existing age changes the answer as life expectancy is based on conditional probabilities. Here's a longevity calculator from the experts who know best: http://www.longevityillustrator.org – Chris W. Rea Mar 27 '18 at 22:40
  • Well, not sure about the US, but in Australia someone born today is expected to live past 100. – Victor Mar 27 '18 at 22:51
  • 24
    OP remember the 25-30 years is not until you need the money, the 25-30 years is until death. You need the money well before that, unless you are trying to maximize your children’s inheritance. You will have no need for the money once you die, but need it whilst your are retired to pay rent, pay for food, play golf and whatever you need to do. When retired you don’t have other income usually, so you need to live of your savings and are thus extremely vulnerable to any decrease (a downward correction could mean that you would then have no money for your last years at all = extreme high risk) – ssn Mar 28 '18 at 00:10
  • 4
    If you have another guaranteed source of income then high risk is fine. But if you need that money to live on you can't afford to risk losing it. You should probably have up to 10 years of safe investments depending on where we are in the market cycle. – kweinert Mar 28 '18 at 00:31
  • @kweinert: Most people will have another source of (more-or-less, depending on politics) income: Social Security ($18K/yr on average). So you need to figure the gap between that, and what you need as a minimum income. – jamesqf Mar 28 '18 at 02:45
  • You have it, risky and while approaching retirement you need to ensure some revenue not taking risk to lose everything because in such a case you will have less chance to restart! Sure to have something acceptable and not risk to have more but also less (isn't that risk?). – Jean-Baptiste Yunès Mar 28 '18 at 10:21
  • 2
    If you end up being old and without money, would it make you happy, that you've maximized the expected value of your retirement income? –  Mar 28 '18 at 14:55
  • What a great bunch of answers. I'm going to have a problem picking one... – Paul Cezanne Mar 28 '18 at 22:40
  • 1
    I would take issue with the "invest more in bonds, for instance" part. Conventional wisdom holds that bonds are safer than stocks, but you should not take this at face value. Especially in the current low-interest environment, bonds carry a lot of risk of their own, and in the last couple of years, bonds seem to have moved more in tandem with stocks than be a bullwark against them. – Kevin Keane Mar 31 '18 at 05:17

16 Answers16

51

What occurred in 2000 and 2008 is "why this is true". If one were to have retired in 2007, with a major portion of their retirement invested in stocks, their portfolio value dropped significantly, possibly by 50% or more - if one is continuing to invest this is not a problem, you get to take advantage of the average dollar cost over your future investments. However if you are pulling money out of your investments to live on, you cannot easily recover the lost values.

Moving a large portion of those funds to more stable investments minimizes such drastic drops in value.

JimmyJames
  • 3,840
  • 11
  • 19
Norm
  • 1,184
  • 6
  • 10
31

Presume that you are using the safe withdrawal rate of 4%. So if your retirement account is $1,000,000, you are withdrawing $40,000 a year.

If there is a market correction, and your retirement account loses 33% of it's value, your options are:

  • Continue drawing $40,000 a year, which is now 6% of your savings
  • Reduce your draw by a third to $26,800

Neither of these is desirable.

On the other hand, it is not expected that if your retirement account gains in value, that your draw will likewise gain; essentially that extra money is just providing a buffer against a correction. So from this point of view, reducing volatility is more important than increasing long term rate of return.

Magua
  • 5,097
  • 1
  • 15
  • 23
  • 2
    While neither is desirable, the first one is correct. This was settled once and for all by a massive institutional rethinking of how endowments work. UPMIFA changed the law from "can draw all gains anytime, but can't draw below its initial value" (which starved charities when they need it the most).... To "can-should draw, even if it's underwater, but no more than 7% even in boom years". – Harper - Reinstate Monica Mar 28 '18 at 20:30
  • 2
    @Harper As is common in real life I don't think either extreme is correct. Should I cut down by a third and risk being unable to afford food or meds or my mortgage. Obviously not. Should I spend 5k on a European vacation? Probably not worth the long term costs and risks of running out of retirement funds. In the end you end up cutting what you can and make sure you cover what you need to. Many people's problem when it comes to money is they don't understand true needs vs wants. – Evan Steinbrenner Mar 28 '18 at 22:58
17

Which would you be more happy with: less money or no money?

The interpretation of the word "risk" in this context is "potential to lose it all", enough reason for anyone to think twice about highly risky investments.

Of course greater risk often brings higher returns, so it's quite literally a game of risk vs. reward.

The deciding factor here is time. Time gives the young the ability to recover from even the harshest of failed investments. If you lose it all at 40, you can still contribute your yearly cap to retirement investments with low risk and have a decent nest egg by 65. If you lose it all at 60, you are, to use the technical term, screwed.

So your best bet is typically, as you say, to make well-informed but riskier investments when you're younger, and then scale down the risk as you approach retirement age. Because having less saved for your retirement is far less unpleasant than having nothing.

TheEnvironmentalist
  • 1,373
  • 3
  • 14
  • 17
  • 1
    Exactly. This points out what people usually don't understand in the word 'risk'. –  Mar 28 '18 at 14:51
  • This answer is so underrated, it addresses the key underlying aspect of the question which is RISK. – NegativeJo Mar 28 '18 at 16:05
  • 2
    This would segway into similar answers about why have Health Insurance... the risk of "bad things" is small - but the damage if it hits you is catastrophic, leading up to bankruptcy and financial ruin. Insurance against a TV at Best Buy? probably a bad deal... Insurance for your life and retirement? Probably a good idea. – WernerCD Mar 28 '18 at 18:01
  • The scenario that you "lose it all" is not really the reason for this. If all the companies in the market go to zero stock value, they won't be making bond payments either. As discussed in other answers the concern is that the value of your stock holdings will drop such that you are forced to sell such a large of a portion of your portfolio to support yourself for the rest of your days. If you are worried that the stock market is going to 0, you should be investing in beans and a bunker. – JimmyJames Mar 30 '18 at 16:56
15

Okay, I'm seeing a lot of answers/comments that hinge on sharp downswings and avoiding them (aka, '2008'). That's not the danger of using S&P or similar for your post-retirement holdings. The danger is a long protracted stall of the fund - such as what happened from 2000-2012. When that happens, you're withdrawing $X each year, for 10+ years, but the fund's not replenishing anything at all. Sure, the economy will probably pick up afterwards, but by that point the damage is done: your account is far smaller than what it started with, and the percentage gains have a much tinier effect.

I did the numbers that Victor uses, starting in the year 2000 (though with the slight difference of not including the year's spending money in the index fund - you can't exactly make interest on stuff you've spent or are about to spend.)

Result? The account is down 50% in 7 years. It was down to $246k by the end of 2011. And sure, the S&P kicked back into gear and started posting double-digit gains... but the problem is, at that point you're pulling 20% of your fund each year. Sure enough, the account is gone by the end of 2018.

(Also, as an aside, you shouldn't figure 7% annual averages from S&P. The average from 2000-2018 is 3.8%.)

That's why it's considered high-risk. You might be fortunate, and the S&P doesn't enter any prolonged stalls in the beginning of your retirement - and your fund might last far longer than if it was in lower-risk accounts. But, you might be unfortunate, and retire just as the economy stalls. And if that happens, your retirement account will deplete much faster.

As to the original question?

I'm not sure anyone advises putting all your retirement money into low-risk accounts. If it helps, think of it this way:

I'm retiring now, with $X in my account. I want to plan on living 30 years. So out of all the money I spend, 1/3 of it will be within 10 years - aka, stuff that needs to be very very stable. Another 1/3 of it will be 10 years down the road - aka, stuff that a temporary market downswing doesn't hurt too much, but a stall will. Another 1/3 of it is 20+ years down the road - aka, stuff that I should still keep in a long-term growth account.

See? It's not a situation with absolutes. You're still keeping some of the retirement fund in non-low-risk accounts. You're just keeping less of it there.

Kevin
  • 2,630
  • 10
  • 16
  • If I was invested in an S&P500 fund in 2000, whether retired or not, I would have pulled out in December 2000 (start of the bear market), I would have been back in September 2003 (start of the bull market), our again January 2008 (start of the bear market), back in again July 2009 (start of the bull market), out in July 2010 and back in September 2010, out August 2011 and back in January 2012, out in January 2016 and back in July 2016. Now if the S&P closes below 2532 at the end of this week or next week I would be out again (as this would be the start of a new bear market). – Victor Mar 29 '18 at 01:06
  • I am no advocate of buy and hold, but my demonstration was to show that just putting all your funds in low risk/low reward just because you are hitting retirement, can be the wrong thing to do because if you end up living longer your money will run out, and that is a risk to be considered. – Victor Mar 29 '18 at 01:12
  • 8
    No, you've got the wrong idea of what risk is. Putting in a stable, reliable fund lets you have a very good idea how long your money is going to last. Putting it in the S&P might last you longer... or it might not.. That's what risk is - you have literally no idea how long your retirement fund is going to last. – Kevin Mar 29 '18 at 01:16
  • 6
    Plus, I think it's really horrible to give the advice of 'Put your retirement savings into the S&P' while silently thinking, 'But you have to be able to predict the markets 10 years out for this to work reliably.' – Kevin Mar 29 '18 at 01:19
  • Firstly, I don't predict any market nor do I try to, I react to what the market is telling me right now and act on it appropriately. Secondly, the risk is not in knowing how long your money is going to last, it is in not knowing how long you are going to live for. If you think that you are now retiring so can put your money into safety and it will comfortably last you the next 20 year, but then 20 years is up and you're still fighting fit but have no money left, and you end up living for another 20 years, then that is a risk you have taken. – Victor Mar 29 '18 at 01:28
  • 2
    @Victor clearly, this strategy does not work. If you were so great at reacting to the market, you would be rich, not writing comments saying "If I was invested in an S&P fund in 2000" but assuming you were still around to write comments at all, they'd read "I read the market and invested in an S&P fund in time to make millions, then exited before a loss". – iheanyi Mar 30 '18 at 20:56
  • @iheanyi - firstly I don't invest in the US market, I invest in the ASX, secondly I only leant what I know over the past 5 years or so - which is why I didn't start investing in 2000, thirdly I have been firstly paper trading, then trading with small accounts, then building it up to larger accounts as I learn more and become more experienced. So I do implement my strategies, and I would not change the strategy just because of my age. By the way I am able to be semi-retired due to my property and share investments doing well. – Victor Mar 30 '18 at 22:07
8

If 2 different people both retired at age 60 at the start of 2008 with $1,000,000 in capital, and one took all their capital out of the stock market and placed the funds in savings earning 2%p.a., whilst the other kept the entire $1M invested in an S&P500 index fund. So who would be better off today and whose funds are likely to last longer?

Have a look at the spreadsheet below:

$1M Capital

Both are taking out $50,000 each year for living off. The one that puts their savings in savings at 2% will run out of money after 25 years, whilst the one that keeps their money invested in the stock market still has $190,000 left after 25 years.

The returns from age 60 to age 70 are the actual S&P500 annual returns between 2008 and 2018. From then on I have assumed another large fall in 2019 of 35% and then from 2020 to 2033 I have assumed the average annual return for the S&P500 of 7%p.a. This is probably a bit conservative because usually after a large fall there are some years straight after of larger than normal annual gains. So in real life it is likely that the second person would have more than $190,000 left at age 85.

This proves that safety is not always the best option, you make the decision yourself.

Victor
  • 20,991
  • 6
  • 47
  • 85
  • 1
    Your data suggests that there is some solid math behind my question. Thanks! – Paul Cezanne Mar 28 '18 at 10:20
  • Not solid, just a simple spreadsheet showing both cases in comparison - the spreadsheet does make it a lot easier than it looks. – Victor Mar 28 '18 at 10:37
  • @Victor I like where you are going with this, but why use actual data for the first 11 years then assumed information for the last 14? Why not just use real data starting at 1993? – Kevin Mar 28 '18 at 12:44
  • 3
    @Kevin - because some of the other answers were highlighting the GFC in 2008 as a reason for running for safety, and I set out to test the situation if 2 people with identical funds went separate ways when they both retired at age 60 in 2008. In 2018 the one invested in the stock market is more than $200K in front. Then I decided to include another market crash followed by average market returns after that. But you can try the same exercise of any 20 or 30 year period - and I am pretty sure you will get similar results. – Victor Mar 28 '18 at 12:56
  • 1
    @Victor, I did. If you go back and "retire" in 1988 at the age of 67 with a million in your fund, you can spend $2M on your 90th Birthday party and still leave your kids $5.8 million when you pass away at the ripe old age of 97. – Kevin Mar 28 '18 at 14:36
  • 1
    Note that S&P 500 index funds are comparatively low-risk as far as stock market investments go (in the sense that if your fund lost all of its value you'd have waay bigger problems than a lack of retirement income). I'm not sure this answers the actual question, because it's not clear what is meant by "less risky". – Jared Smith Mar 28 '18 at 14:40
  • 5
    I cant downvote, but this is wrong on so many levels : a) returns for safe investment are not accurate (2% over the long run, at least from an historical point of view is understated). b) future returns are maybe correct from an average perspective but not from a standard deviation perspective. Volatility has an impact on the chance of having no money before your end of life. This is risk , would you take a 5 or 10% chance of running out of money even if it meant leaving 200k more to your children ? – NegativeJo Mar 28 '18 at 15:52
  • 1
    Ha, I can't upvote more than once. If you have a lot invested then the downturns don't affect you, if you don't have much invested then you need it to continue to grow while you are retired. – AbraCadaver Mar 28 '18 at 15:55
  • 13
    Running the simulation with historical 40 years of returns and inflation data, leaves you with 5x more chance of running out of money in 10 years if you are in stock than in AAA bonds. This is risk. I can give you an infinite number of "games" that are worth playing (from an average or expected value persepctive) it doesnt mean you can afford to play them from a risk perspective. This is why as you get older you should shift to a safer portfolio (to answer the original question -- and not to go into a perma bull, stock fanbois rant) – NegativeJo Mar 28 '18 at 16:02
  • 13
    All this proves is that 7% flat return returns more than 2% flat return. Additionally, this is why you adjust your allocations as you age, no reasonable person suggests pulling out of the market 100% and allocating 100% to a 2% APY savings account. – quid Mar 28 '18 at 17:45
  • 4
    I'd hate to see people follow this advice and keep all their retirement savings invested even in S&P 500 right up until the end when there are a variety of stock & bonds rebalancing techniques that will outperform both of these investment techniques. – zzzzBov Mar 28 '18 at 17:49
  • 18
    Uh, this is terrible, terrible advice. Starting at 2008 for this sort of analysis is horrible. Sure, that year starts out with a sharp downturn, but that's not where the danger is - the danger is a long plateau, where you're regularly taking out money each year but your fund isn't getting anything to replenish it. I did the same spreadsheet, starting in 2000. The S&P puttered awhile, not shrinking a whole lot, but not growing a whole lot either. End result? That million is gone in 18 years - because the S&P spent 11 years in a stall. – Kevin Mar 28 '18 at 18:18
  • Why do you expect a crash in 2019? Economy is great again. – Harper - Reinstate Monica Mar 28 '18 at 20:36
  • Well there could be one this year, the S&P500 is not far away from officially being in a bear market, if it closes below 2532. And I was trying to be conservative with the investment option. – Victor Mar 28 '18 at 21:51
  • @Kevin -this is not advice it is a demonstration, showing that the normal way most people think of just running for safety when they hit retirement is wrong. Personally I would be in the stock market no matter if I am in retirement or not, as long as there is a bull market. But as soon as the market changed to a bear market I would pull out of the stock market. Whether I was in retirement or not would not change the way I would do things. Plus I wouldn't and don't have all my funds just in the stock market, I would have some in property and some in cash, just as I do now. – Victor Mar 29 '18 at 01:18
  • 10
    @Victor You've filled out almost 60% of the time period with a flat 7.0%. The only thing you've demonstrated is that advice based on avoiding large and unpredictable fluctuations doesn't make sense if you take the large and unpredictable fluctuations out. –  Mar 29 '18 at 08:14
  • @ChrisH - and to be fair I have added a -35% fall prior to the average of 7% annual returns. Usually the 7% average includes the rises and the falls, but to be on the conservative side I have included one large fall then only average rises - so realistically after a large fall you would have had some years of larger than average returns (which I didn't include). Also as mentioned, this answer is not advice, it is a demonstration to show that just putting you funds in low risk/low return just because you are in retirement is not good advice and can actually be risky. – Victor Mar 29 '18 at 09:09
  • 8
    I did a quick python simulation to check out this advice. For each year I picked a random year from the last 30 years of the S&P. I then repeated 100000 times. While most of the time the stock market did better I noted how often the money ran out early (aka DISASTER/DESTITUTE). 0.87% of the time within 10 years, 2.6% of the time within 20 years, 5.6% of the time within 30 years. To compare, at 4% interest it lasts 42 years and at 5% interest it lasts forever. – Mark Perryman Mar 29 '18 at 12:08
  • @MarkPerryman - so your saying that 90.93% of the time the money lasted longer than 30 years. Also, at interest rates of 4% and 5% inflation would be quite high, so you would also need to be taking out a lot more than $50K per year after 10, 20 or 30+ years to still be able to maintain the same living standards. So taking this into consideration those funds might last shorter periods from which you stated. – Victor Mar 29 '18 at 12:30
  • 5
    It is cumulative: 94.4% of the time it lasted longer than 30 years. Inflation works equally well on the stocks so is kind of irrelevant. The key is that a 5.6% chance of being old and penniless is not a risk to ignore. So people reduce the risk, even at the cost of average pot size. – Mark Perryman Mar 29 '18 at 12:38
  • So where can you get a cash rate of 4% to 5% today? I thought the 2% I used was on the high side. Also, as I have mentioned, this is not my advice, I don't recommend to buy and hold, I recommend risk management with any investment. This was just an illustration, and you have just added to that demonstration. Plus I wouldn't recommend having all your funds in the stock market at any time. But what I do recommend is to not just run to safety just because you have retired, some growth assets are still required when you are in retirement just like during your working life. – Victor Mar 29 '18 at 12:46
  • 8
    At age 72, the person earning 2%/year should have moved over to the 100% reliable 7% per year investment you gave the other party! – Yakk Mar 29 '18 at 16:01
  • 1
    DV because this is a very misleading example. Your S&P retiree was 15% down when you ran out of real data and assumed a market return of 7% for everything thereafter. – J... Mar 29 '18 at 19:25
  • @Yakk - If it was me personally, and started in an S&P500 fund in 2000, whether retired or not, I would have pulled out in December 2000 (start of the bear market), I would have been back in September 2003 (start of the bull market), our again January 2008 (start of bear market), back in again July 2009 (start of bull market), out in July 2010 and back in September 2010, out August 2011 and back in January 2012, out in January 2016 and back in July 2016. Now if the S&P closes below 2532 at the end of this week or next week I would be out again (as this would be the start of a new bear market). – Victor Mar 30 '18 at 00:06
  • 3
    @Victor Losses hit harder than gains, especially during drawdown. Seriously, without some sort of intelligent stochastic modelling this argument isn't worth much. The right way to do this is to run computational simulations for all kinds of scenarios and run an analysis of the statistics. A half scenario for one real time-period is an anecdote, it isn't science. The 2% return for the conservative investor is also a bit weak. You can do better than 2% with a conservative portfolio. A serious analysis would examine the risk spectrum a bit more thoroughly. – J... Mar 30 '18 at 08:22
8

Actually, finance theory doesn't suggest a general relationship between one's age or proximity to retirement and the riskiness of one's portfolio. According to modern portfolio theory, the riskiness of your portfolio should be related (only) to your risk aversion/apetite.

The "common knowledge" you mention comes from the fact that many people become more risk averse as they age, either because they rationally fear not having enough money to last to the end of their lives, or simply because their personality changes in their old age.

However, as you point out, there could be a large amount of time between retirement and death so retirement is not a good target date. Moreover, people's preferences and wealth differ. A wealthy person who is not worried about having enough to live on but would like to maximize the expected amount they leave to their heirs, for example, would likely want to increase their risk as they age.

Bottom line: "Common knowledge" only applies to some people. You should let your risk aversion dictate the riskiness of your portfolio, irrespective of your stage in life.

farnsy
  • 15,050
  • 30
  • 51
7

It's not so much that younger people have longer to ride out market swings, it is that they are continuing to save.

If a thirty year old loses a third of their savings, it's bad, but not only with that third most likely recover, but also it is not the whole of their retirement savings - their best earning (and hopefully saving) years are ahead of them.

Rupert Morrish
  • 7,566
  • 4
  • 26
  • 41
6

For decades, the general rule was an allocation ratio of 100 minus your age for equities which decreased your market exposure as you grew older. The remaining portion was placed in safer assets such as investment grade bonds.

Until last year (an anomaly?), life expectancy has been increasing. Coupled with historic lows interest rates, many have questioned whether this rule is still practical. Some investment advisers are now recommending using 110 and even 120 instead of the traditional 100 minus your age so that along with longevity, we have additional years of portfolio growth.

It really isn't this simple because there are other factors to consider: size of retirement portfolio, other sources of income, lifestyle and expenses, even gender.

Despite what I have written, I don't follow this at all. After many years of margin trading, I'm now far more risk averse and I am at a much lower equity exposure than 100 minus my age because I am confident that my assets will last and keeping them is far more important to me than making it - though I don't object to making it. So AFAIC, the short answer is that it depends on your individual circumstances.

Bob Baerker
  • 76,304
  • 15
  • 99
  • 175
  • 1
    That's right because someone with a much lower pool of money at retirement might need to take on risk even in retirement. So maybe some risk management is also needed. – Victor Mar 27 '18 at 22:49
  • I'm liking the simple formula, 110, or 120, minus your age. Easy to remember and not a bad first approximation. – Paul Cezanne Mar 27 '18 at 23:21
5

UK based answer: Until recently, on retirement you were forced to buy an annuity with your pension pot by age 75 (ref: saga website). This meant that if the market dipped just before you were about to buy your annuity, and didn't recover before you turned 75, you were forced to buy an annuity with a much reduced pot.

The rules have now changed. You don't have to buy an annuity with your pot, and if you do decide to there is no longer an age limit.

The advice for people approaching retirement age to de-risk is therefore now less important than it used to be. But not unimportant (see other answers).

AndyT
  • 821
  • 6
  • 10
  • But if you don't buy an annuity, the alternatives are to take the money as a lump sum (and if you take too much you'll be taxed on it) or to wait (and you need money to live). So you may be able to take a tax-free lump sum rather than buy an annuity just after a dip, and live off that, but if recovery takes longer than you guessed, you'll still be buying an annuity in a bad patch – Chris H Mar 28 '18 at 15:24
  • 1
    @ChrisH - Yes, you're completely right. This is why I have said that de-risking is still not unimportant. But there's no longer a point at which you're forced to buy an annuity, which is why it's now less important. – AndyT Mar 28 '18 at 15:31
5

I think I got some interesting results with Monte Carlo simulations.

TL;DR: Balance your portfolio.

Many people answering this question worrying about a big drop right after retirement, mentioning 2008. So I ran my simulation like that.

  • Initial amount: 1,000,000 USD
  • Withdrawal amount: 45,000 USD / year (inflation adjusted each year)
  • Start year: 2007
  • End year: 2017
  • Rebalancing annually

Here are some results.

  • US Large Cap 100%: Link US Large Cap 100% 2007-2017
  • Total US Bond Market 100%: Link Total US Bond Market 100% 2007-2017
  • 50%/50%: Link 50%/50% 2007-2017

It seems many answers are concerning sudden drops right after retirement. But this may misdirect many people to go safe investments(like bonds) all the way. I think retirement fund should be seen in the long term, rather than focusing on recent one big drop.

Look at the charts I showed you. I believe retirement funds should last at least 30 years, and the 50%/50% portfolio has the highest success rate. Possibly you can tweak more to get better outcomes, but you get the idea. Unbalanced portfolios are not good for the long-term, hence not good for retirement portfolios. Research yourself for the better options, but don't waste your time tweaking too little things.

Edit 1: As @Pace said, Monte Carlo simulation has its flaws. So I tried backtesting this time(Link). The basic setup is somewhat similar:

  • Initial amount: 1,000,000 USD
  • Withdrawal amount: 45,000 USD / year (inflation adjusted each year)
  • Start year: 2007
  • End year: 2018
  • Rebalancing annually

Backtest 2007-2018

In the result, Portfolio Growth (inflation adjusted) graph shows 100% bond portfolio's value is gradually decreasing during the period, faster every year. The other portfolios definitely took the loss during 2008 but recovered afterward. Although the loss is huge, if retirement portfolio last at least 30 years, this is not the major issue.

Here is another test(Link). Backtest 2007-2018 with Long Term treasury

This time, I used Long Term treasury instead of Total US Bond Market. Interestingly, in the recovery period, both US Large Cap 100% and 50%/50% portfolios are increasing almost in parallel. 50%/50% portfolio also has highest Sharpe ratio and Sortino ratio too.

While backtesting also has flaws, my point still stands. Balanced portfolios will perform well in both good and bad times. In the long-term, of course.

Edit 2: If we change backtests' start year to 2000 as @kevin suggested in another post(Link), Total US Bond Market 100% is the all-time highest. Backtest 2000-2017

You see, that's why I don't like backtesting. It's useful for a certain case study, but it makes hard to generalize strategizes. It makes us very biased. Also, nobody knows what will happen in the future. Focusing on a case might not be a good idea for a long run. Be mindful when backtesting.

BTW, we can change our strategy when our situation is changed. 30 years is a long time. Be flexible. It's your money after all.

Edit 3: Let's try Monte Carlo tests starting 2000 this time(US Large Cap 100%(Link), Total US Bond Market 100%(Link), 50%/50%(Link)). Total US Bond Market 100% has the highest success rate for 30 years. But be careful. If you are expecting your retirement portfolio last more than that(like 35, 40 years), you will get very different results. you can check this from the links' Portfolio Success graphs.

  • US Large Cap 100%'s success rate starts dropping early on compared to the others. But the success rate is relatively high in the late years. US Large Cap 100% 2000-2017

  • Total US Bond Market 100% is doing very good for 30 years. But after that, all scenarios are starting failing rapidly. Maybe you can prevent this by changing to Long Term Treasury, but that will lose its risk-averse purpose; 50%/50% portfolio with Long Term Treasury has a less standard deviation in this case. Total US Bond Market 100% 2000-2017

  • 50%/50% doesn't look better for 30 years since Total US Bond Market 100% is doing well. But after that, its success rate basically remains the best. 50%/50% 2000-2017

user2652379
  • 715
  • 4
  • 13
  • Is portfolio visualizer your product? – quid Mar 28 '18 at 23:42
  • @quid No. Did I make a mistake? Should I mention this in the post? – user2652379 Mar 28 '18 at 23:45
  • I think it might worth mentioning that changing Total US Bond Market to Long-Term Treasury from 50/50 portfolio increased success rate to above 99%. – user2652379 Mar 29 '18 at 02:05
  • That isn't simulating the crash in the way you think it is. It computes the mean return across the entire period (+9.82%/year) and the standard deviation (14.64%) and then simulates each year by randomly sampling from that. This will not cause a re-enactment of a crash. Even at the 50th percentile there is a positive return each year. – Pace Mar 29 '18 at 04:16
  • @Pace For that case, I think you can use 10th Percentile line in Simulated Portfolio Balances graph. That basically represents worst case scenario. Of course, it's not 100% real-life simulation, but still good for adjusting our strategies. – user2652379 Mar 29 '18 at 04:18
  • @Pace To be fair, nobody knows what will happen. Monte Carlo simulations are good for seeing many possible scenarios. I think we should prepare for the best case as much as the worst case. – user2652379 Mar 29 '18 at 04:28
  • @Pace I added backtest results in the post, simulating the actual historical crash. – user2652379 Mar 29 '18 at 11:59
  • That looks great! I think this approach still suffers from some of the critiques against Victor's approach (note: if you change the start year to 2000 as @kevin suggested then US bonds win) but the visualization is much better than a spreadsheet :) – Pace Mar 29 '18 at 14:46
  • Can you do screencaps (legally) to embed the images here? Answers shouldn't heavily depend on external links to make sense. – Yakk Apr 02 '18 at 17:35
  • @Yakk I not sure if that's legal. I'll ask them – user2652379 Apr 03 '18 at 02:57
  • @Yakk I added screenshots as it is legal. This post becomes too long though... – user2652379 Apr 03 '18 at 04:35
  • @user2652379 I don't agree that "we should prepare for the best case as much as the worst case". The first priority has to be assessing the risk of the various worst-case scenarios. The second priority has to be ensuring that you will survive any reasonably likely worst-case scenario financially. Only when you have covered that should you even start to look at the best-case and the most likely case. I also see a problem with your diagram because when you are 60, a 30-year time horizon is much longer than your life expectancy. – Kevin Keane Jun 30 '18 at 00:15
  • @KevinKeane I agree that I was wrong about the first one. About the life expectancy, according to SSA from USA(Link), A man reaching age 65 today can expect to live, on average, until age 84.3. ... And those are just averages. About one out of every four 65-year-olds today will live past age 90, and one out of 10 will live past age 95.(I'm not American though) – user2652379 Jun 30 '18 at 03:37
  • @user2652379 Good point about the life expectancy. I believe one common way to treat living past age 85 or so is to consider this one of the risks (i.e., it should be a factor in your worst-case scenario). But you are right, 65 year olds do have a life expectancy of around 20 more years. – Kevin Keane Jul 02 '18 at 15:38
4

The assumed goal behind the recommendation is that you don't want to run out of money before you die. If you want to leave money for heirs or maximize your retirement income that changes the goal and recommendations. Generally 4% withdrawal is considered "safe" based on the trinity study from the late 90s. If your investments simply keep up with inflation you have 25 years at 4%. So you don't need to make a high percentage on your money in order, but you do need to avoid loosing your investment.

The first couple of years of retirement are far and away the most risky for retirement. If you retired in 2007 with 1 million and planned on withdrawing 4%, if the market drops 50% like it did in 2008, now to maintain your lifestyle you withdraw 40k of 500k a full 8% of your savings. Very few market scenarios would allow you to recover from this situation short of going back to work or dramatically cutting your lifestyle. If however you only have 40% in equities suddenly you have 800k and a a much better chance to recover.

I am not sure the advice is as helpful if you have some mitigation strategies in place for the first few years. Either going back to work part time or cutting expenses dramatically can lessen the impact of the loss significantly.

stoj
  • 4,699
  • 17
  • 24
4

The short version is this:

The common practice you speak of is built on the premise you will be withdrawing principal and depleting the fund in retirement.

You move to less risk, lower reward funds because you are no longer contributing to their growth. If you saved correctly, you don't need them to grow, you need them to last. The last thing you want is for a swing in the market to take an unexpected chunk out of your principal. You can more safely budget your withdrawals if there is a much lower risk of a force outside your withdrawals taking money out of the fund.

Also, you need it to last FOR YOU. If you want something to be left in the fund after you die, you would need to find that balance of risk and return outside of this general practice, which is intended to prevent losses, rather than make gains.

cornbread
  • 165
  • 4
3

Most people of retirement age have decided money in excess of their basic needs does not make them happier. A Florida trailer park and a public beach are just fine, but but going back to work after being out of the job market for a decade is not an option.

So there's some amount of income which is "enough", and the objective is to find the cheapest portfolio which is almost certain to last until death. Let's say "almost certain" means there's a 5% chance of being wrong.

Consider a hypothetical conservative portfolio, with an expected mean annual return of 4%. In any given decade the returns may be more or less, but in 95% of the decades the annualized return is at least 3%.

A more aggressive portfolio may have a mean return of 8%, and in 95% of decades the return is at least 1%.

For retirement planning, it's not the mean return (4% vs. 8%) that matters, but the "worst case" return (3% vs 1%). If the objective is to have a portfolio that is 90% likely to not run dry, the conservative portfolio is cheaper.

As the time horizon broadens the worst case become less bad, but the horizon must be very broad, several decades, before an aggressive portfolio's worst case is better than a conservative worst case. Risk comes not only from sharp market crashes, but also protracted periods of poor performance. Consider the Nikkei 225 index (Japan), which hit a high at the end of 1989, and has been up and down within the same band several times since.

One must be very young for such a broad horizon to exist.

Phil Frost
  • 3,131
  • 15
  • 17
1

In addition to the consideration of risk related to age, there is another, related, consideration. Risk has been discussed in many other answers already.

As you retire, it is important to shift from growth to income-producing investments. This is easiest explained by example of somebody who chooses to invest in real estate for his retirement security. For somebody in his 30s and 40s, it makes a lot of sense to buy houses, or build them, even if you have to take out a mortgage for it. You are building your wealth here. When you are in your 70s, you want to have a consolidated portfolio that gives you regular income from rent checks. You are not using your wealth, but rather you are reaping the fruit of it. And if you have enough buildings, on occasion you can sell of one to splurge on something in retirement.

At first glance, it may appear that stocks are different because they are passive rather than active investments. But they really aren't all that different. During your 30s and 40s, you'd buy growth-oriented stocks, even some high-risk stocks (like Yahoo, Pets.com or Google in the dotcom era, or Tesla, Uber etc. today), and hope that they have matured into solid businesses by the time you reach 65.

Of course, some people (or their funds) trade stocks much more actively, but that doesn't necessarily produce better results.

Personally, my favorite investment when you retire? Pay off your mortgage. Even if you have to sell off most of your 401(k). Paying off your mortgage gives you a guaranteed return of 4% or more depending on your interest rate, in the form of interest-free and rent-free living, for the rest of your life.

Kevin Keane
  • 537
  • 2
  • 5
-1

Actually it's true for everyone, the 'common knowledge' is wrong. The best strategy, for all ages is 90% in some inflation hedged cash equivalent, and 10% in ultra volatile very high risk, but potential high reward investments. This is called the barbell strategy. See the works by Nassim Taleb explaining this. The main books where this is explained from an investment perspective are "Fooled by randomness" and "The Black Swan".

Don't believe this? See this article which also shows that, even over very long periods, just switching your money between the best savings accounts available can beat the stock market, but with no risk to your capital. This doesn't discuss the barbell strategy, but does show that the excess returns from the stock market over virtually any period less than about 20 years do not actually exist because of occasional big falls, and there is a risk that at any particular time, you could be unable to withdraw your capital because you have to wait 10 years for the market to bounce back from a big fall. From the article:

Savings accounts beat the total returns on the tracker in 57% of the 192 five-year investment periods beginning each month from 1 January 1995 to the present. The tracker won in just 43% of periods.

Looking at longer time periods, the results were even more marked. For example, over the 84 14-year periods from 1995, cash beat shares a whopping 96% of the time.

Looking at a range of investment periods since 1995, from one to 11 years, the analysis found investments in funds that track the FTSE 100 would have actually lost money up to a third of the time. By comparison, there's no risk if you put cash into a savings account – you always end up with more than you started with.

The other part of the strategy which is to invest 10% in more risky things would involve things like out-of-the money put options, or things like an unlisted business with high potential for growth etc. There's no easy answer to this part unfortunately.

Note this answer has been been downvoted, but no-one has actually refuted my points.

Community
  • 1
crobar
  • 115
  • 2
  • 3
    That is so counter intuitive. I might need to research this more. But a savings account paying ~1% should never beat "the market" which is paying ~7%. Now, if you dive more into it, this was in the UK and it seems that savings accounts are tax-free, so that would certainly help. And I'm not in the UK... – Paul Cezanne Mar 28 '18 at 10:25
  • 1
    @crobar Can you point to some research or works that show examples of this? What are the "ultra volatile very high risk" investments that are recommended? – stoj Mar 28 '18 at 12:10
  • The risk is loosing that 10%. Well, no, actually it isn't even a risk, because you're almost sure to loose that 10%. The only thing that you 'get' is the expected value (buy a dog, you'll have 3 legs on average). –  Mar 28 '18 at 14:54
  • 2
    The linked article in the second paragraph doesn't support the absurd claim in the first paragraph, or even connect to it in any logical way. The article basically says that from 1995 to 2016, a British index fund averaged a 6% return, while "best-buy savings accounts" were returning 5%. –  Mar 28 '18 at 17:33
  • Re Nessim Taleb, he has made a specialty of studying "black swan" events, i.e., extreme statistical fluctuations, some of which may occur because statistical distributions may have "fatter tails" than is usually assumed. It is true that a high-risk investment may have a high expected return, and some finite probability of extremely high returns. But if it's also got a 90% probability of being worthless in five years, it's probably not the magic key to your long-term financial well-being. –  Mar 28 '18 at 20:26
  • @BenCrowell, You would need to read Nassim Taleb's books, particularly fooled by randomness, for the evidence. Further evidence, the investment in high risk ventures is exactly the strategy the big VC firms employ. They expect 99 out of 100 to do little or go bust but the remaining 1 in 100 make enough of a return to offset this. Investing in the 'medium risk' (whatever that means) stock market, there is a small chance on any given day that you might lose a huge proportion of your capital (see the 1920s market crash). That might be ok, if you have another 30 yrs to wait for it to come back. – crobar Mar 29 '18 at 08:49
  • @BenCrowell, To make profits, one must first survive, that is the strategy. To survive you need to insulate yourself from large movements. Large but rare movements are a characteristic of the stock market (the 'fat tails' you mention). Most of the advice to invest in the stock market comes from people wanting to sell you some form of investment. – crobar Mar 29 '18 at 08:53
  • @BenCrowell, and the article actually states that over virtually any period less than 20 years, cash (in the UK in an ISA, which is a tax free vehicle you can put a certain amount in each year) beat or almost matched the stock market with absolutely no risk of losing any of your money! – crobar Mar 29 '18 at 08:56
  • @9ilsdx9rvj0lo, but this is exactly the strategy venture capital employs, invest in high risk things, and 1 in 100 makes enough of a return to offset the loss of the rest. Plus, at the end, you still have 90% of your capital for sure. – crobar Mar 29 '18 at 09:03
  • @stoj, I've added some content to the answer to address your comment – crobar Mar 29 '18 at 09:14
  • Savings accounts seldom pay as much as inflation, so you are losing spending power the whole way. People keep seeking tricky ways to come out ahead. In the end, braodly diversified stock market investments have seldom done poorly over the long-run. Unless you are going to focus on the Depression, where the governements did everything possible wrong, there is not really a more clever strategy. And taxes matter a whole lot. – eSurfsnake Mar 29 '18 at 18:32
  • low volatility and long-term investments tied to the growth of the economy - index funds - will do the most over the long-run. Bar-bell strategies, savings account rotation, and other ideas are just nonsense. – eSurfsnake Mar 30 '18 at 04:17
  • This is a cherry-picked result designed to generate clicks on savings account referral links that pay the publisher commission. Consider "the 84 14-year periods from 1995". "14 years" is an odd number to pick, isn't it? And wasn't there a famous bubble in the 90's? These periods fall into three categories: 1) buying on the upside of the 90's bubble, but exiting just after the 2008 crash. 2) buying at the peak of the 90's bubble. 3) buying on the downside (but not the bottom) of the bubble, and exiting in the low from Aug 2015 to Jun 2016. – Phil Frost Mar 30 '18 at 11:15
  • @PhilFrost, No it's when records of best buy accounts started. Plus he looked at more than just the 14 year period, did you read the article? It beat it over virtually every period he looked at. He looked at every one of the periods chosen every month from 1995.Also the guy who did the research is well known in the uk, he presents a money show on BBC radio 4 (a national station). Moneysavingexpert is also a very well know site in the UK and is known for being impartial. There's a stock market crash roughly every decade. One before would have been 1987 etc. – crobar Mar 30 '18 at 18:20
  • "The observation was correct for virtually every period he looked at" is the definition of a cherry-picked result. – Phil Frost Mar 30 '18 at 19:41
  • @eSurfsnake I can't vouch for his strategy (and I'm skeptical), but I suspect that if it works, it may be because the return from stocks is overall not as high as most people assume. We usually look at average returns, but a relatively few people (such as venture capitalists, who shoulder extraordinary risks in pursuit of high rewards) profit disproportionately from the stock market. Average investors, almost by definition, will see substantially less than the average return. – Kevin Keane Mar 31 '18 at 05:27
  • @KevinKeane I think average investors by definition see exactly the average return. There's no doubt this article is a misleading scheme to generate referral revenue. Don't lend it credibility. – Phil Frost Apr 03 '18 at 12:22
  • @PhilFrost No, that's actually the fallacy. OK, it does depend on the definition of "average investor". In the strictly technical sense, you are right - but that average investor simply does not exist. The problem is that with any average, some people will make more and some will make less than the average. We know who makes more: people like Bill Gates, Elon Musk, Warren Buffet and more.Their extraordinary returns have to be balanced by somebody's portfolio underperforming. It's mathematically impossible for "everybody else" to, collectively, even match the average. Agreed on the misleading. – Kevin Keane Apr 04 '18 at 04:52
  • @PhilFrost, the average return for the stock market is actually not computable from only a sample of the data from all time. 'Average' and 'mean' calculations based on a sample of the whole of time assume a certain statistical distribution. For the stock market, for which the distribution of prices is a power law (rather than, say, a uniform or normal distribution), the mean cannot be calculated from a sample because the errors are huge. So saying what the average return from the stock market is without knowing what happens over all of time is not actually meaningful. – crobar Apr 04 '18 at 07:16
  • @eSurfsnake, that's why I stated inflation hedged cash equivalents in my answer. I only provided the link to show that long term stock market returns are inflated, and even a very basic strategy with cash can beat them. – crobar Apr 04 '18 at 07:27
-3

The common wisdom is very wrong.

People used to retire at age 65 with a pension from a large company and some savings. 50 years ago, if you lived to 65, your expected lifespan was less than 5 years. Today if you live to 65 in the U.S., your expected lifespan is to the mid-80's. And, of course, some people will live longer. So you need to plan to age 95 to be safe.

Today, few people have a pension - they have their own savings, likely in a 401|(k). Let's suppose that is the case. |And let's suppose you have saved $600,000.

The wrong way to think about this is 'let me move it all to bonds and live off the interest'. Why? (1) you would only get, at best, 3% interest. That's only $18,000 per year, and (2) taxes will be very high on bond income.

But its worse. If it is currently in stock you bought long ago, you would need to sell that first to buy bonds. Suppose you have a 25% tax rate; if it is a 401(k) maybe 33%. SO, right away you have reduced your $600,000 to $400,000 to $450,000. At 3%, that pays $12,000 to $13,000 per year.

Really, over the long-run - and 30 years is the long run - the stock market is a better bet. People worry too much (like in the post above) about a recurrence of 2008. But, that is a very rare occurrence. In reality, the market goes up on average about 8-10% annually. And, dividends are about 2% as well. So, $600,000 will pay you about $12,000 in dividends or so per year.

Plus, on average, the stock goes up $48,000 to $60,000 per year on top of that. So you can spend some of those profits as well (which, like dividends, will be taxed at 15%, or 20% at most - unless in a 401(k). Even if markets are anemic, or you are conservative, you can spend likely $30,000 plus the $12,000 of dividends - $42,000 - and at that rate the account would keep going up in value as well.

People worry way, way too much about a market meltdown. If you have 30 years to plan for, you will live through one or two. But unless you have so little money that a 25% drop in a single year that recovers over the next 4 years (sort of what happened), stocks are a much better investment.

Here is a great old rule. Whatever percent you ear, divide 72 by that number and that is how long it takes to double your money. Even if you get 7%, in 10 years your account doubles. If that happens, and a 30% market drop happens, you are still way ahead.

People focus way too much on the short-run and buy things like bonds. In many cases the only certainty they get, if they knew how to calculate it, is a certainty of going bankrupt before dying.

eSurfsnake
  • 217
  • 1
  • 3
  • 1
    Taxes will be very high on bond income? 401(k) Distributions are distributions regardless. Why on earth would anyone cash their entire 401(k) immediately upon retirement to right away reduce your $600,000 to $400,000? And the stock market does not go up 8-10% on average PLUS dividends; it's provably closer to 7 (including dividends). The common wisdom is fine it's your understanding of taxes and distributions that's very wrong. – quid Mar 29 '18 at 18:05
  • I pulled the data myself recently. From 1950 to 2017, the S&P 500 had total returns of 11.5% with about 17% standard deviation. That includes the stagflation years of the late 70s and early 80s. There are a bunch of chicken little who run around saying stock returns are about 7%; they are going down; bonds are close; etc. None of it matches the data. And here is another issue: a 60/40 stock/bond fund, over 30 years, delivers about 30% of stock market returns. Where these ideas come from, with no data to back them up, escapes me. And, believe it or not, most people don't have a 401(k). – eSurfsnake Mar 30 '18 at 04:15
  • Ok, January 1950 to January 2017 your number is right, your answer says over 30 years not 67. Care to comment on: (2) taxes will be very high on bond income. or SO, right away you have reduced your $600,000 to $400,000 – quid Mar 30 '18 at 05:06
  • 1
    Re: "In reality, the market goes up on average about 8-10% annually". It won't continue for long at these prices. Future returns are highly dependent on valuation levels. Those kinds of returns are more likely from market troughs, not peaks. – Chris W. Rea Mar 30 '18 at 13:25
  • The 30 years was related to the 60/40 portfolio, not market returns. On other comments, in general stocks as a whole - assuming long-term constant P/E multiples - is the sum of (1) population growth (more people buy more stuff); (2) productivity growth; (3) inflation (prices and costs go up, but so do profits, all else equal, in dollars), and (4) corporate leverage (at least 1:1:1 for the S&P 500 in total). if the first three numbers are 2%, 2%, and 2% you get 6% EBITDA growth. A little leverage gets you 12% long-term growth. Short-term valuations are noise; this is the 50 year view. – eSurfsnake Apr 05 '18 at 04:22
  • Your analysis of the tax impact is way off the mark. As long as you keep the investments in a 401(k) or roll it into a traditional IRA, the tax is exactly zero regardless of whether you move from stocks to bonds or to cash or to any other form of investment (of course other transaction costs do apply). You only pay taxes on whatever you actually withdraw, or when you convert to a Roth. And if your annual withdrawals are as low as you say, your income may end up being low enough so even that could be tax free. – Kevin Keane Jun 30 '18 at 00:01
  • yes, in a 401(k). – eSurfsnake Jul 04 '18 at 13:14