On Yesterday’s Bloomberg Surveillance show, Tom Keene referenced some of my bullet points from Keep Investing Simple. (One bullet was to Ignore the Chattering Class, but I was not referring to Tom, whom I have a great deal of respect for).

Live TV being what it is, we briefly touched upon rebalancing, but gave it short shrift. A few emails from viewers and a comment from Econochat later, it is clear that I muddied the waters rather than clarifying them. This post will hopefully correct that misadventure.

Whenever I reference rebalancing, I am referring to asset classes — Equities, Bonds, Real Estate, Commodities, etc. It does NOT refer to individual stocks.

Take a portfolio model of 60% equities 40% bonds (60/40) primarily held through broad ETFs. After there was a hypothetical big quarter for bonds and a weak quarter for stocks, your 60/40 portfolio ends up looking more like 58% Equities and 42% bonds. The market action has led your holdings to drift away from your allocation. Rebalancing means that you are returning back to your original asset allocation model weightings.

Why do this? Because history teaches us that all broad asset classes will eventually mean revert. The goal of the rebalance is to take advantage of the short term volatility and price swings to move you back towards your model 60/40 allocations at more advantageous valuations.

In practice, it means you trim your bonds holdings after they had a good run and you buy equities after they have fallen. If you have more asset classes, you do the same, rebalancing to your model. A typical allocation may be that are 62% equities 31% bonds, 4% Real estate and 3% commodities 62/31/4/3. On a regular basis, quarterly, semi annually or annually, you rebalance back to your original allocation.

50/50 is a very conservative allocation, 60/40 more  moderate, 70/30 more aggressive, 80/20 even more so. The original allocation decision you make is a function of your risk tolerance, assets, and time horizon.

If you own Berkshire or Visa or JNJ (as we do) and they are working, you lock them in your portfolio and let them run. With individual stocks, you DO NOT rebalance — but eventually, you may have to make a sell decision. That is an entirely different conversation.

Asset classes mean revert. Any stock can go to zero, but an asset class cannot. And the reason to own an individual name (versus an index) is that it has the potential to outperform that broad index. That’s why you DO NOT rebalance them — it defeats the primary purpose of owning individual companies. .

Category: Investing

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

18 Responses to “A Word About Portfolio Rebalancing”

  1. [...] Barry on portfolio rebalancing across asset classes versus stocks.  (TBP) [...]

  2. AHodge says:

    i dont trade as long and forget
    but if i did
    i would rebalance classes like you say and recommend that for more sleepy non-traders
    you make an excellent-and for me brand new
    argument to not balance the individual names.

    this does raise the question–what names do you buy on a downturn? then you are into stock picking?
    either you have a special picked porfolio or a generic etf type
    but if you just rebalance the whole thing
    you are in effect rebalancing each of the names in it?
    i agree you can make a decision not to rebalance names you love on the upside-let em run
    but if you have a generic etf that is real messy for that part
    would have to sell ETF buy the names separately?

  3. constantnormal says:

    Barry, you (I think) tend to look askance (putting it nicely) at “buy-and-hold” … when you refer to a bond segment of a “balanced” portfolio, are you thinking of these bonds as things to be traded, scampering up & down the yield-maturity curve, capturing imbalances in value and yield, or as “buy-and-hold” investments in that asset class?

    Same question with regard to owning property … are you more favorable/averse toward flipping, renting, buy-and-holding, or are these all equally valid choices, and a matter of personal suitability?

  4. gloeschi says:

    • I agree that in normal times an ‘agnostic’ portfolio serves you best (you don’t take any bets on market direction, because market will mean-revert)
    • Agree with portfolio weights (bonds/stocks) based on time horizon (1 year -> can’t own stocks, 20 years -> high weight in stocks). So, for most of us, it would be a function of age (the older you are, the lower the weighting in stocks).
    • I would add the following: Buying standard off-the-shelf market ETF’s is cost-efficient, but not performance efficient (since, if they track a major index, the ETF’s will be market-capitalization weighted, which makes you automatically invest most heavily in those stocks currently in fashion). Many studies have shown that if you use ANY other method of weighting (equal weight, weight by revenue, weight by first letter of the company, etc) you OUTPERFORM (since you avoid to be biased towards stocks that are temporarily en vogue. Think Deutsche Telecom at EUR 100 (now EUR 9) or Research In Motion at $140 (now $12)). This is, by the way, the reason why Astrologers seem to have success in investing; it’s basically investing by random, and random beats most mutual funds.
    • Heavy bond-weighting in older age exposes you to elevated inflation, or, worst case, currency reform. You mostly own paper-money. In extremis, an 80-20 portfolio of bonds-stocks will lose you 80%. Keep that in mind.

  5. These accounts tend to be 401ks, IRAs, long term retirement planning and the like.

    As to Buy & Hold vs timing, we allocate a portion (e.g. 25%) to a tactical portfolio we work with that is either all equity or or bonds based on a quant model that changes the investing posture monthly.

  6. Lord says:

    Another technique is to set allocation windows for each asset class and when an allocation hits one bound, reallocate to the other bound. This reduces reallocation frequency while allowing a measure of allocation against the trend.

  7. tippet523 says:

    We automatically re balance when a position becomes double weighted. IE 5% goes to 10% any time before a normal rebalancing. It did allow us to trim a lot of Apple at close to peak prices. I do hold agree to hold and let winners run but not when they become to large a percentage of the portfolio. Great Piece

  8. ShakyShot says:

    Why Quarterly? does it give the best return, or just meets the (questionably sound) objective of maintaining a fixed allocation over time? Aren’t asset price cycles considerably longer and therefore a longer time would give higher return? Why not a percentage divergence trigger? Indeed, why one of these arbitrary trigger at all, but rather use some data that indicates an assets run might be over?

    gloeschi Says:
    “Buying standard off-the-shelf market ETF’s is cost-efficient, but not performance efficient (since, if they track a major index, the ETF’s will be market-capitalization weighted”
    True, the most used ETF’s are Cap-weighted, but equal and other weightings are now available in good ETFs for slightly higher cost.

  9. Rightline says:

    Tom Keene gave you a big shout out on this post at about 9:10am on Bloomberg radio. He used glowing words like; clinic, tour de force, and must read.

  10. [...] light of this morning’s portfolio rebalancing discussion, I thought the table above would help you to visualize the [...]

  11. [...] rebalancing applies to asset classes, not individual stocks. (Big Picture, [...]

  12. CANDollar says:

    William Bernstein advises rebalancing your index portfolio less frequently than yearly – even every several years.

    Why? Because the longer time frame better captures market cycles and less frequent rebalancing will reduce the impact of trading costs and possibly taxation – frictions which over time add up to a lot that compound to lower returns.

  13. Frilton Miedman says:

    Jumped here from the “Asset Class Returns” blog.

    Interesting graph, often the biggest loser one year is among the biggest winners the next.

    Seems a large scale variation of the Dogs of Dow idea based on the longer time frame of the Kitchin (business/inventory) cycle would make it a good idea to re-balance according to one to two year performance.

    Over longer secular time frames, wage growth, disparity, employment rate, consumer income to debt ratio’s seem to reflect market direction & momentum,

    The way outliers like Kyle Bass & Michael Burry deduced a major correction was coming back in 2006 when by noting RE prices were substantially outpacing wage growth seems to apply to the bigger picture.

  14. Rich in NJ says:

    How do you decide between letting a winning position run and taking profits?

  15. [...] is also a typical time that investors look to rebalance their portfolios. Barry Ritholtz at the Big Picture asks (and answers) why investors should have a plan in place to rebalance their portfolios. He [...]

  16. courageandmoney says:


    How would the classic Portfolio look like with ETF? SPY, qqq, IYR, HYG, TLT?



  17. gregskidmore says:

    It has been a few years, but I remember reading in one of David Swensen’s books how Yale will rebalance some of its equity portfolios daily. I believe he siad it provides them an additional 1% in returns.

    On a separate note I’ve concluded that rebalancing between assets of similar risk (standard deviation), say commodities and US stocks, in theory takes advantage of mean reversion without changing the risk of a portfolio. Rebalancing between assets of different risk, like US stocks and US treasuries, provides a mean reversion benefit and and ensures the risk in a portfolio returns to its intended state. One is performance rebalance and the other is a risk rebalance.

    I’ve found in backtesting that a performance rebalance can be done quite frequently, daily, weekly, monthly, quarterly…. a risk rebalance needs to be done infrequently, like once every one or two years.

    I discovered in back testing risk rebalancing through the great depression that you can rebalance yourself out of money. Selling bonds and buying stocks quarterly through a period like that can have devastating results.

  18. bonzo says:

    For every dollar of positive alpha, there must be a dollar of negative alpha. Mean reversion can only work if there is some dummy out there doing the opposite of mean reversion, like chasing performance. And there are such dummies, we know that. So why not target these dummies directly rather than using a indirect approach like mean reversion. That is, when the dummies get euphoric and greedy, it’s time for the smart money to get fearful and take money off the table. Don’t need mean reversion for that. Just look at the valuation and sentiment metrics, plus some “gut-feel” sentiment metrics if you have one of those magic guts.

    Selling at the top is a lot harder than buying at the bottom. This is because, fear is palpable at the bottom, whereas tops tend to drag on forever and end out with a whimper rather than a bang. Right now is a case in point. We’ve been stuck at slightly under 1500 on the SP500 since September. I sold back in October myself. Would have sold in September but I needed to wait to get the long-term capital gains rate on stuff I had bought back in Sept/Oct 2011. Thank god I don’t have a blog of my own or a public persona or I’d have to be listening constantly to the pea-brains with the attention span of a chipmunk mocking me for having sold too early. I’ll have my comeupance eventually and this won’t be the first time I got out a little early and missed the last few points of a blowoff. Year 2000 was particularly painful, as I recall, because it wasn’t just a few points back then and the waiting for vindication took forever. If I could hold on through year 2000, I can certainly hold on now.