Tuesday, March 03, 2015

Passive management vs. saving more

So, Eduardo Porter wrote an article, with large excerpts from John Bogle, telling people they should switch to passive management. Here's an excerpt:
On average, a typical working family in the anteroom of retirement — headed by somebody 55 to 64 years old — has only about $104,000 in retirement savings, according to the Federal Reserve’s Survey of Consumer Finances...The standard prescription is that Americans should put more money aside in investments. The recommendation, however, glosses over a critical driver of unpreparedness: Wall Street is bleeding savers dry [with fees for active management].
Ryan Decker does not like this. Bogle and Porter, he says, are "straightening the deck chairs on the Titanic". The real problem, he says, is not that people are paying too much in fees, but that they aren't saving enough in the first place. Ryan writes:
So people are approaching retirement with $104,000, and "a greater part of the problem is the failure of investors to earn their fair share of market returns."...Do Porter and Bogle really want us to believe that the main reason people are trying to retire on $100 grand is that they haven't made sufficient use of passive funds?... 
I'm as big a fan of passive management as anybody, but this is totally absurd...What would the average nest egg be if everyone had chosen the right fee structure and asset allocation? Whatever it is, it's not going to get anyone very far in retirement, particularly if they're accustomed to spending money at the kind of rates that lead to having such a small stash at that age... 
By all means, don't throw your money away on active management, and don't waste your time trying to pick stocks. That stuff matters at the margin. But that particular margin is insignificant compared to the problem of low savings rates.
Ryan is right, and he's also wrong.

Ryan is right that saving more of your income is capable of supporting a lot more retirement spending than saving money on management fees.

But Ryan is wrong to say that this makes Bogle and Porter's argument invalid. Actually, Bogle and Porter have a good argument. Here is why:

If you consume less today in order to consume more during retirement, you have to give something up (today's consumption). Thus, Bogle and Porter are recommending something that gets you more total consumption, while Ryan is recommending something whose main effect is just to move around your total consumption. Reducing fees is no substitute for saving more when it comes to increasing retirement consumption. But saving more is no substitute for reducing fees when it comes to increasing lifetime consumption.

Let's make that concrete. Suppose I save $1000 more when I'm 30 and consume $1000 more when I'm 65. And suppose my per-period utility is u(c_t), and the discount rate over 35 years is B. Then my utility gain from saving more is u(c_30 - 1000) - u(c_30) + B*u(c_65 + 1000) - B*u(c_65), where c_30 and c_65 were the amounts I would have consumed, before I decided to save more. The first part, u(c_30 - 1000) - u(c_30), is going to be negative. The second part, B*u(c_65 + 1000) - B*u(c_65), is going to be positive. In other words, the utility gain from saving more is only the utility gain from more consumption smoothing. And if people are already doing the optimal amount of consumption smoothing, that utility gain is zero.

Just for fun, let's think about that point a little more deeply. We should ask Ryan: Why do we think people aren't doing the optimal amount of consumption smoothing already? What sort of behavioral bias is keeping them from making the right choices for themselves? Do they have hyperbolic discounting? Do they have self-control problems? What's wrong with these people? If nothing is wrong, then saving more would actually be bad for them.

(The same, of course, could be true of active management. Maybe people are choosing active management, and paying the fees, because they get some utility out of doing so. It's certainly possible. Of course, if you take that position, you should then explain why more and more people are switching to passive management, just as Bogle has been urging. Preference shift? Seems unlikely.)

But anyway, the point is that you can't simply compare saving more with paying fewer fees in terms of the increased retirement spending the two are capable of supporting. That's like ridiculing people for picking up a $10 bill by pointing out that they could get a lot more than $10 if they sold their car.


  1. If 2.27% 'effective fees' is correct (per NYT) and we take the S&P 500 as our investment, that means starting with $33 a month in February 1980 and investing $28324.58 total.

    2.27% fee estimate:
    XIRR: 7.43%
    Final Value 02/2015: $104,111.70

    1.12% fee estimate:
    XIRR: 9.215%
    Final Value: $150,400.84

    0.06% fee estimate:
    XIRR: 10.441%
    Final Value: $195,420.04

    Assumptions: I did the math on the S&P 500 using Bogle's numbers, assuming a 60 year old who started 35 years ago and increased his monthly investment by inflation. I used roughly $104,000 as a final balance and the 2.27% 'effective expense' ratio suggested in the Times article, along with a 1.12% and a 0.06% fee scenario.

  2. I would like to pose this question on every economic blog to see whether anyone can give a coherent answer to it:

    If every person saved the optimal amount for retirement in the optimal ways, what would the average retirement income be?

    I am willing to bet my entire derivative portfolio that nobody can answer that question.

    And so at a time when overall consumer demand is too low to grow the economy at the desired rate, what is all this talk about people not saving enough? Clearly most people don't have enough saved for retirement, but my question is this: is it possible for everyone to save enough for retirement as most financial advisors would advice?

    Of course not.

    Please, let's get real.

    1. "And so at a time when overall consumer demand is too low to grow the economy at the desired rate..."

      There is no particular level of "consumer demand" necessary to "grow the economy." Keynes certainly would not have fallen for this nonsense: what is needed is a level of *spending*, which is consumer spending + investment spending.

  3. With my above comment, I am certainly suggesting that there's a huge problem in expectations caused by a financial advisory business that is pretty immoral. Yes, the fees gobble up much of the savings, but on top of that, they advertise nonstop that we all need to invest more and more.

    Ah, there's no reality in this. I know this industry from the inside to some extent. It isn't necessarily totally corrupt, it is just a bunch of people going to work trying to make money and nobody is looking over their shoulders to make sure it is moral, equivocal, feasible or even reasonable.

    That is all it is.

    1. How about the most basic optimal savings guidepost - "income net of taxes and savings"?

      If you believe the Trinity study/4% number (aka, don't look at Japan), multiply that number by 25 to get a starting point.

      Sure, there are a lot of problems with that simple target (tax treatment of assets, house payment or lack thereof, moving to a more expensive/less expensive place, etc, etc) - but in terms of a "most of the people most of the time" it isn't too bad, in my mind.

    2. Ok, I'll look into this, but given your calculations applied to the US, what is the answer?

    3. I'll play: ~$270,000 for a median household.

      That's actually a high target though, because a 60 year old retiring in 6 years is probably closer to paying off the house than most. (He or she can also do a reverse mortgage). Also, with an 8% savings rate that shouldn't be an issue - but if you want I can run some backtests.

      The math is easy but tedious - maybe I'll make a calculator based on this, haha. You can actually get a reasonable estimate for most people doing this...

      Assumptions/work: Median household income: let's round to $54,000, one earner, 1 exemptions at state and federal level (60 year olds usually don't have kids in the house, and I'm assuming standard deduction due to long-ago home purchase). I arbitrarily picked Virginia as a middle tax state. Paycheck City gives me: $39,528.72 net.

      My original research for a typical net savings rate at the $54,000 level (split the difference on this calculator: http://dqydj.net/how-much-do-people-save-by-income/ ) is 8.03%. That leaves us with $36,354.56

      As for Social Security, I used Bankrate's estimator. Age: 60, retire at 66, raise 1% a year inflation 2%. It estimated $25,967 a year in Social Security, meaning we need to replace $10,387.56 in income a year. Multiply it by 25 (or 33 for a 3% withdrawal rate - your risk tolerance matters) and voila, this household should have $259,689.09 saved. Now round it off to $270,000 because of 6 years of 1% raises and me not wanting to estimate tax rates in 2021.

  4. "If you consume less today in order to consume more during retirement, you have to give something up (today's consumption). Thus, Bogle and Porter are recommending something that gets you more total consumption, while Ryan is recommending something whose main effect is just to move around your total consumption. Reducing fees is no substitute for saving more when it comes to increasing retirement consumption. But saving more is no substitute for reducing fees when it comes to increasing lifetime consumption."

    But Noah, it's all about utility, not about total lifetime consumption. That's a potential means, or a factor, but what you care about is maximizing total utility. Now, I know that people criticize the use of utility in economics. But the problem is in measuring and defining it. Saying you want to maximize this thing called utility is just saying you want the best deal out of life with regard to what's important to you. This is very sensible and a good way to think about things and evaluate which bundle, or strategy, or path is best.

    Now, there are caveats. Mentally ill people may have very self-destructive desires, or utility functions. And utility is powerful as a way to find what is best, but is not necessarily that predictive with regard to how people actually behave in an economy.

    But it's really important to determine, or estimate well, what is the best option, even if people don't take it. So, I'm a big fan of the utility framework, even though I recognize it's crucial to do it well and interpret it intelligently.

    That said, do people save a lot less typically than the utility optimizing amount? I think yes. And this is a bigger factor than getting an average of 6% real in the stock market rather than 6.5% real, about the historical average.

    But it is complicated. Positional externalities are enormous and profound. Drive a Civic (like Noah!) when your reference group drives Lexus's, and $45,000 minivans, and huge SUV's, and your wife is like, no one of our income level does that. What's wrong with you? You're so cheap.

    By the way, my wife is extremely good with personal finance, and largely in agreement with me, but most spouse's wouldn't be.

    And, I love Civics' and have owned them; I had an Si I loved. But we have always had cars way less expensive than typical for our income and wealth levels.

    But a big point is positional externalities. And very impacted by them, having to attract and keep a good spouse can really make it hard to save the otherwise utility optimizing amount.

    Finally (for now), with managed funds it's not just the fees at all. These managers, to sell their skills, will diversify a lot less than the Vanguard Total Stock Market Index. They'll make bets on sectors, on industries, even single stocks. So the average active fund is a lot riskier, especially to laypeople who often have no idea how much risk they're taking.

    Finally (part II), having substantial money for retirement is for so much more than petting turtles in the Galapagos and constantly touring the country in your $200,000 RV. It's about not living in destitution in your 60s, when age discrimination is off the charts. It's about being able to afford giant co-pays on crucial medical treatment. It’s about not being a terrible burden to your children, often struggling in a far more dangerous world.

    1. Reading on, you do go with the utility optimization framework! Hey, it's comments, so I might informally go off on something before reading the whole. Often it can be that or nothing, with time constraints. Sorry.

      I'll have to look at that more closely. But I will say, in the real world, perfect consumption smoothing is not optimal for most people. Low money consumption will hurt you a lot less as a young and healthy man than as an old one, or a middle age one with children depending on you. Plus people feel a lot better when their consumption rises over time than when it's stagnant, or disastrously declining (like a self-positional externality), and there's more, and academic support. I talked about this is a post once that got a response from Conor Clarke when he was with the Atlantic. See:


    2. Another interesting point, that I'd really like to explore more one day. What about one of thee managed funds that gives you pause --- Berkshire Hathaway! Freakin' Warren managing your money!

      My first though is I would love to just hand a big chunk of my money to him and his team with just the right philosophy. But the question is at what price? How much more are those Berkshire Hathaway component assets costing you because Warren is a rock star in the finance world? Are these fans pushing the price of $300 billion in underlying assets up to $400 billion because it’s Warren? Is it worth it? I did some research, and owned some BH in the past, but I just wasn’t comfortable enough with this issue, and let’s face it, Warren’s probably too old now to be doing much of this kind of intensive work. How good is the current team?

      But here’s a very recent analysis that interestingly shows a small premium for the BH management:


    3. I'll add, and I think this is very importantly, Noah asked once, what do you do about positional externalites – as if this is hard – make a rule against gold jewelry, something like that was done in the past, and rich people then made their buttons of gold, and started adding buttons to their coat sleeves, which is something we still have.

      But as I've said, the answer is really pretty easy and general, just highly progressive taxes with the proceeds paying for largely non-positional goods.

      And here is a great example, if you increase taxes progressively to pay for doubling social security payments, now retirement gets far more secure, but what gets cut in return? When people are younger, and if doing well, making a high income, they have less, so $50,000 Lexus instead of $100,000 Mercedes; or Honda Accord EX instead of a Lexus; but since the utility difference is predominantly positional, few utils are lost. A new Honda Accord EX with leather and all of the ding-dongs will give just about as much pleasure if it's just as exclusive and prestigious as new Lexus was, because everyone in that reference group is spending less on cars due to the higher Social Security tax. So, for little loss in utils, now everyone's retirement is far more secure and far less stressful, and far less often tragic.

  5. Not mentioned is the most important problem of all: people do not make enough money to retire, no matter what. Inequality again rears its ugly heard.

    Second, taking the 2.27% fee as given, hiring a manager reduces gross expected annual returns by the fraction of the fee divided by expected gross returns. The lower the expectation, the greater the take.

    Now here's the magic part. Such a deal is equivalent to forking over to the manager that fraction of the total *principal* up front in exchange for zero fee thereafter.

    Thus if expected returns are maybe 6%, the transfer to the manager is roughly 1/3 of principal. I merely suggest that is obscene, and a measure of the casual oligopoly that exists. Consider the aggregate implications, that the money management industry effectively owns 1/3 of all the savings.

    I'm in the biz, so this is a really big deal.

    1. Now here's the magic part. Such a deal is equivalent to forking ...

      You left out the effect of compounding. Over long horizons the numbers are worse than you suggest.

      Investors, and advisors, tend to underestimate the time periods involved. I made my first retirement account contribution when I was thirty. I expect my wife (then widow) to withdraw the last of the money more than sixty years later.

    2. I get it; high fees are bad, but we don't need to resort to ridiculous examples to demonstrate this. The notion that 1/3 of principal is transferred is silly. And where are averaging Americans finding all these investments with 6% expected returns combined with 2+% in fees?

    3. Six percent real returns seems to be about what the stock market has returned in the past. I'm not sure where the 2% came from but I note that 2% is where the hedge fund industry starts when setting fees so 2% is not out of line.

      Suppose we take slightly different assumptions: thirty year holding period; tax free account; 4.5% total real return; 1.5% management fee. One dollar invested with the management fee gives you $2.43 and without the management fee gives you $3.75. Management fees matter a lot. The fact that most active mutual funds under perform the market just makes the case for passive investing even more compelling.

      Yes, under these assumptions most of the burden of retirement will have to fall on an increased savings rate, but switching to passive management and cutting fees is the low hanging fruit.

  6. Decker is right. His critique is spot on; your's is all but useless. Decker explicits recommends passive management. He's not saying don't pick up the $10. He's saying if you need $1 million in 20 years and you currently have nothing, you're main focus should be saving more. And please, spare us the BS about the completely rational average American retirement saver. It's ridiculous. I'd argue that for a large portion of the people in question, the goal is to shift consumption from early years to retirement, not about maximizing total consumption.

    1. Please ignore the numerous grammatical and spelling errors in my post. But I made them and know they are there, so it must have been a rational decision.

    2. you're main focus should be saving more.

      If I want to save more I have to work more and/or consume less. On the other hand simply switching from actively managed accounts to a passive investment in, say, the S&P 500 index can cut my management fee by enough to be equivalent to a doubling of my contribution if we are looking at a thirty year investment period. Add in the fact that most money managers under perform the index over the long haul and passive becomes pretty compelling.

      One way of looking at your retirement account (to a first approximation) is that you contribute from age 30 to 59 and withdraw from 60 to 89. Then the money you withdraw at 60 is the money you put in at 30, plus returns, and the money you withdraw at 61 is the money you put in at 31, plus returns, and so on. So even with the money you put in at 60 you have a thirty year investment period.

      My son is twenty four and he is about to make his first investment. It is going straight into a broad stock market index ETF.

    3. So even with the money you put in at 60

      Should read:

      So even with the money you put in at 59.

    4. Yeah I get all that, as does Decker, and we're in agreement that the majority would be much better of in passively managed funds. But active management fees certainly aren't "the greater part of the problem" as Bogle says. Decker's critique was spot on. Perhaps his analogy was too strong though because rearranging the deck chairs on the Titantic would have done no good whereas a switch to lower fees will do some good. If a guy goes to the doctor and reports he's been smoking 4 packs of cigarettes a day for 20 years and also has a cigar every Saturday night but wonders why he has lung cancer, does the doctor tell him the that darn weekly cigar was the problem?

    5. Most people have what non-economists call "the zero lower bound problem". They have to pay money to eat. They have to pay money for a place to live. They have to pay money for transportation to get to their job to get any money at all. They have to pay for health care. If they don't make their basic nut, they start what is called "deflation" which ends with them living in the street, eating at soup kitchens and so on. Any savings they have is basically a buffer against the car needing repair work, an illness, a layoff or one of the thousand shocks that flesh is heir to.

      In other words, they are making rational decisions to avoid reaching the zero lower bound.

      If there is any money left over, they save it for retirement. As mainstream economists like to say, "Good luck with that."

    6. Derivs2:50 PM

      "I'd argue that for a large portion of the people in question, the goal is to shift consumption from early years to retirement, not about maximizing total consumption."

      I would agree. When I was in college I made a reverse spreadsheet. I entered what I want saved by 65 and then went about filling it in. I figured at a young age one can have a lot of fun and make sacrifices easier. At an older age, it is not a time one wants to make sacrifices. Much like the philosophy of taking risk at a young age because you can always make it back. I just wanted to make sure I would never be "without" at an age I couldn't take the risk.
      What I found was that saving early allowed me a great peace of mind as my equity built. It allowed me at a young age to buy more expensive cars that depreciated less and actually cost less than cheaper cars, options became available to me that exponentially shortened the time necessary to achieve those goals.
      Now I tell my nieces and nephews how to do the same. Never did think about maximizing overall consumption. Simply future safety.
      My other rule. Try not to buy too many depreciating assets.

  7. Anonymous9:37 AM

    I am reaching that age where the loss of function is becoming noticeable on almost a monthly rate. So I would recommend to the young to screw retirement and just get on with life that is available when young.

    1. Anonymous9:23 PM

      Good to keep in mind. That may be one reason folks don't save more. On a very micro-level, I broke my leg last month. I acutely feel the passing of every weekend in which I'm unable to take a bike ride or a mountain hike or just a long walk with the dogs. No amount of spending on these weekends cooped up indoors could make up for that. It occurs to me with your comparison that every weekend I wasted on working, even though it boosted my retirement account, is still trading fun today for money that won't be much fun to spend when I'm old.

    2. This is a good example of how the "zero lower bound" works in real life.

  8. Anonymous2:41 PM

    "Preference shift? Seems unlikely."

    How else do you explain it?

    1. Reputational effect slowly overcoming severe information asymmetry about probabilities of long-term outcomes.

  9. Anonymous8:13 PM

    I don't see myself as being someone who will care as much about consumption when I'm 65 or 70 as I do now as a guy in his mid-30s. So why be overly worried about saving for 30 years from now? Plus who knows I may die before I hit 60, who the hell knows. I think the 30s are probably the best decade to match most of the fun consumption with. You're old enough that you have some seasoning in life yet still young enough usually to still look and feel young still.

  10. Anonymous9:46 AM

    Where is it written in "The Book of Life" that you can work 25 years then take the rest of your life off.......?
    Some earn enough, some do not....

  11. Anonymous9:54 AM

    I think Noah failed to comprehend Decker - I don't think Decker ever said Bogle's argument was invalid.

  12. Anonymous10:02 AM

    At the age of 68 - these days make enough to break even on expenses. Haven't tapped retirement savings yet, but taking SS now and putting half of it (i.e., half of 40k) into personal retirement account.

    From mid '80s to 2000's working in software development making in six figures lived on 1/3, paid 1/3 in taxes, saved 1/3.

    Being a drop-out in the 60's and seeing the poverty side of life, cannot recommend leaving things to chance when your ability to earn is over. Plan for your old age - it could be scary.

  13. Anonymous11:16 AM

    Eat, Drink and be Merry is a pretty short term world view. If you want to spend 100% of net income on lifestyle during your earning years, be prepared to live on SS from a government that is well know for its ability to wisely handle money. Putting 15% of annual income in tax defered diversified passive fund should keep you off the streets in retirement and not kill all of your party needs when young.

  14. Anonymous11:29 AM

    Why are you assuming that you are only able to earn enough to exactly balance inflation with your extra savings? If the argument was "save $1k now so you can spend $1k later" then of course, that does actually assume that the utility function is non-linear in a way that compensates for the discount factor (that is, "you will optimize your utility by not starving when you're old but you are being stupid and discounting age-65 utility at too high a rate").

    But is that the argument? The S&P returned something like 10x your investment over the last 35 years adjusted for inflation. So you are in fact increasing your total consumption by saving more, assuming that you are investing savings in something you intend to grow in real terms.

  15. Anonymous11:31 AM

    Which gains you more? Saving twice as much, or not paying 1/2 of your retirement savings in fees and lost compounding? Both would be ideal, but if you're choosing between buying your fund manager a BMW and buying one for yourself, the choice is easy. (my own choice is a Subaru and achieving six figures in net worth on a four figure salary).

  16. Anonymous4:14 PM

    If you're young with stable employment and health, then time is on your side. Yet a large and growing % of the investing population that includes retirees and workers near retirement are "underfunded" in their plans. What may be impossible, is trying to "obtain that much money" through conventional methods, such as the "buy and hold" of an index or a changing mix of stock indexes and bond indexes adjusted by a process derived from Modern Portfolio Theory; products offered by the big retail firms and the "new" Robo advisories. The industry doesn't explain that a 7 - 8% (nominal, not adj for inflation) return on these options, accompanied with attendant risk, won't lift these workers very far out of their underfunded state (and maintain the lifestyle to which many are accustomed) with the limited pre-retirement investing years that they still have.
    ... An innovative approach which has produced a higher trajectory of growth with risk mitigation may offer a solution ... https://docs.google.com/presentation/d/1Ua-R53o7c588nUr705hY4YaBEcG1qOrNvtonQIwpVws/edit?usp=sharing

  17. Say they invest the money in SPY, wouldn't the real formula be u(c_30 - 1000) - u(c_30) + B*u(c_65 + 1000*(SPY_65/SPY_30)) - B*u(c_65)? Bogle's point is that people don't maximize (SPY_65/SPY_30). Decker's point is that people don't include the (SPY_65/SPY_30) factor in the decision process at all and if they did, then they would be saving more. Maximizing that factor should be second fiddle to considering it at all.

  18. Anonymous6:55 PM

    Hear hear, Noah. You are 100% correct that its not an either or debate. Improve savings rates, improve investment returns. we gotta to both.

    We blogged about this too. See: http://blog.stockflare.com/blog/2015/03/16/cutting-fees/

    Looks like we all read the same report from the Center for American Progress. :)