Pension Drawdown 3.3 Income Harvesting - Alternatives to Periodic Portfolio Rebalancing to Increase Retirement Income

Image of a crop growing and a bag of money

Whilst periodic portfolio rebalancing with fixed equity and bond ratios is accepted custom there are many alternative strategies promoted that can increase the drawdown income from a retirement portfolio. 

Periodic Portfolio Rebalancing 

The generally accepted wisdom is that portfolios should be periodically rebalanced to bring each asset class back to its original target percentages.  There are three reasons why this is considered important:-

Risk Management:   Portfolios are designed to maximize growth subject to a defined risk.  The risk is usually defined in terms of a portfolio´s volatility,  maximum anticipated drawdown (the fall in value when stock markets decline), or the sustainability of the portfolio in drawdown. 

The volatility of a portfolio is usually reduced through the addition of uncorrelated assets such as government bonds.   The higher the bond component of a portfolio the more stable it will be but at the cost of growth potential.   This is clearly illustrated by the data from Vanguard
in the table below:-

Table of portfolio growth and volatility versus bond content

If a portfolio isn´t periodically rebalanced then as stocks grow faster than bonds as time passes the equity portion of the portfolio will dominate the portfolio and the "riskiness"  of the portfolio will increase and it would be quite possible for a retiree to have a near 100% stock portfolio towards the end of his retirement.

Keeping the Portfolio Design Intact :   Most portfolios have been meticulously researched, designed and back tested in order to have determined characteristics.  Maintaining the balance between the constituents is fundamental to maintaining these characteristics so periodic rebalancing is essential.  For instance, Vanguard´s 60% Equity LifeStrategy fund contains 17  index ETFs/funds and even though some only have a 0.7% representation in the fund it has to be assumed that these percentages are fundamental to the design of the portfolio.

Overcoming Human Instinct to Sell High and Buy Low: private investors are often accused of buying during market bubbles and selling during the crashes -  buying high and selling low. Periodic rebalancing forces the investor to behave against his instincts and sell assets when they are rising and buy assets when they are falling.  

Truth or Myth - Rebalancing Reduces Portfolio Growth?:

I remember in my early days of investing it was said that rebalancing improved the growth of a portfolio by around 0.5%. I was certain this was from Vanguard but haven´t been able to find the reference. However, just using the data from the earlier table this performance enhancement is borne out.  Bonds are shown to have a 6.1% average return and equities 10.3%.  A 60% equity portfolio with no balancing should have an 8.56% average return but Vanguard´s balanced portfolio has 9.1% a 0.44% improvement.  This improvement seems logical but the consensus amongst "experts" is that more benefit is obtained by riding from the momentum of rising markets rather than selling into them.

Portfolio Visualizer allows you to simulate portfolios with or without periodic balancing.    The table below compares the growth rates with and without annual rebalancing from nine different (USA-based) portfolios.  You can see the exact portfolio constituents on the Portfolio Visualizer website.

The results certainly don´t support the "momentum" case against periodic balancing with 8 out of the 9 portfolios producing slightly higher growth when annually rebalanced.  

Table comparing the growth rates of 9 portfolios with and without annual rebalancing

Balancing or Not For The UK?  There is a little more evidence that for UK biased portfolios that there is a slight improvement in performance with no rebalancing.  Analysing 81 thirty-year investment periods starting in each year from 1910 to 1990 with an initial 60% UK equities (basically FTSE All share) and 40% gilts (duration 15+ years).  

54 (81%) of the periods showed improved performance when there was NO rebalancing. Not a great improvement, the average was 0.37%, but over a 30 year investment period, this represents a healthy increase in the portfolio value.  The 27 periods that benefited from annual rebalancing showed an average of a 0.4% gain in annualised growth.

Not Rebalancing Creates Risk or Does It?  Of course, the consequence of not balancing is that the equity portion of the portfolio will start to dominate and this will produce more risk and more volatility.  Of the 81 portfolios, fifty had an equity content of more than 70% at the end of 30 years and thirteen had more than 90%.  

In the next section, I´ll look at alternatives to periodic rebalancing, sometimes called Income Harvesting, some of which have been shown to improve the Safe Withdrawal Rate (SWR) of a portfolio in drawdown despite many of these strategies increasing the equity portion of the portfolio during drawdown which is counter-intuitive to the normal advice of reducing bond holdings with age.

Alternatives to Periodic Rebalancing - Income-Harvesting

Income Harvesting strategies are sets of rules that decide which assets of a portfolio are sold in order to minimise the damage withdrawals make to a portfolio.  In times of market downturns, there is certainly illogicality in pursuing rigid portfolio rebalancing.  If the equity component of a portfolio falls by 50% but bonds increase by 10% the sensible course of action would be to sell bonds to produce the drawdown income and then rebalance.  This is one possible Income Harvesting Strategy which can simply be defined as:-
  • Sell the best performing asset to generate drawdown cash
  • Rebalance
 Clearly, investors in multi-asset products such as Vanguard´s LifeStratagy have to accept the daily rebalancing inherent in these products.

Is Portfolio Rebalancing Necessary in Drawdown?

An early examination of portfolio rebalancing was undertaken by Spitzer and Singh in their paper
Is Rebalancing a Portfolio During Retirement Necessary?.  They examined 6 bond equity allocations,  a range of withdrawal rates, and 5 different withdrawal strategies over periods of up to 30 years.  Withdrawal strategies:

  1.  Withdraw money from either stocks or bonds and then rebalance the portfolio annually to the initial stock/bond proportion. This harvesting rule will be referred to as “Rebalance.”
  2. Withdraw money from the asset that had the highest return during the year and do not rebalance. This will be referred to as “High First.”
  3. Withdraw money from the asset that had the lowest return during the year and do not rebalance. This will be referred to as “Low First.”
  4. Take withdrawals from bonds first and do not rebalance ("Bonds First").
  5. Take withdrawals from stocks first and do not rebalance ("Stocks First)".

Spitzer and Singh concluded:-

" bonds first, over stocks, the best of all the methods, though the resulting stock-heavy portfolio may make some investors uneasy. This method also is most apt to leave a larger remaining balance at the end of 30 years, while rebalancing leaves the smallest amount.

The table below shows clearly the far greater final portfolio value achieved through a Bonds First strategy.  Annual Rebalancing performs poorly.

Table showing final portfolio balances for different withdrawal strategies
( table taken directly from Spitzer and Singh)

One of the most comprehensive reviews of the most common strategies can be found in Michael McClung´s excellent book "Living Off Your Money". He examines 14 common strategies and compares these with annual rebalancing.  Details of these strategies can be found online but later I will describe one of these in detail. 

The table below shows the SWR, average final portfolio value and average minimum bond holding for 13 income-harvesting strategies in the USA, UK, and Japan with the strategies ranked on the basis of the maximum safe withdrawal rate (SWR) with 90% certainty of success.  A degree of caution is needed when reviewing these results as although here there is shown to be consistency between the performance of each strategy across the three countries this was not the case when he undertook the same analyses with SWR-100 (100% chance of success).

Table comparing income-harvesting by country

The table below shows the performance of different income-harvesting strategies for the UK market with an SWR based upon 100% success.:-

Table showing the performance of different income-harvesting strategies in the UK market

The tables show that Income-harvesting strategies can significantly improve the SWR compared with annual rebalancing but at the cost of increasing the volatility of the portfolio over time as the most successful strategies significantly deplete the portfolio´s bond holding. Many retirees would be unhappy with having a 90% or so equity content in their portfolio maybe midway through their retirement when conventional wisdom advises an increase in bold holdings with age.  Having said that, it's interesting to note that one of the most successful and simplest strategies is "Bonds First" which in the case of the UK not only provides a high final portfolio value but an SWR only 0.1% lower than the best.

McClung analyses each method in depth and finally concludes that the best strategy overall for US investors is "Prime", closely followed by "Alternate Prime" both providing significantly higher SWRs than Annual Rebalancing.  Bonds First works well for the UK investor and is probably a preferable option due to its similar SWR, high final portfolio balance. and simplicity

Here is how Prime Harvesting works:

1.  Each year the investor must take the value of his initial equity investment and increase it by the rate of inflation over the period.
2.    If the value of the equity portion of the portfolio is more than 20% above the inflation-adjusted initial equity value the sell 20% of the equity holding.
3.  If there is sufficient cash from any equity sale use this to fund the annual pension drawdown income if there is insufficient cash sell bonds.
4.    Use any cash balance to buy more bonds.
5.     Rebalance if necessary within the bond and equity holdings


Initial portfolio $100,000 50/%50 %bond/equity. 3.5% initial SWR.

Year 5 

Most recent consumer price index: 135.5
Consumer price index at the start of drawdown: 112.5
Inflation-adjusted drawdown income: $4215

(IE) Inflation adjusted initial equity value 135.5/112.5 x $50,000 = $60,222
(AE) Actual Equity Value:    $68,123
Equity value compared to inflation adjusted initial value (AE-IE)/IE x 100 = 13.1%

Equity value only 13.1% above inflation-adjusted initial value so sell bonds to fund drawdown income.

Year 8

Most recent consumer price index: 168.2
Consumer price index at the start of drawdown: 112.5
Inflation-adjusted drawdown income: $5233

(IE) Inflation adjusted initial equity value 168.2/112.5 x $50,000 = $74,755
(AE) Actual Equity Value:    $97929
Equity value compared to inflation adjusted initial value (AE-IE)/IE x 100 = 31.0%
Equity value 31% above inflation adjusted initial value so sell 20% of the equities:_
20% x $97,929 = $19,585. Deduct drawdown income of $5233. 

Purchase bonds with the cash of $14,352.

The process is exceptionally easy and armed with the most recent inflation index and portfolio valuation the calculation and broker operations should take less than 30 minutes per year.


Certainly, there is evidence that Annual Rebalancing doesn´t produce the optimum results in drawdown and that other strategies can allow a higher withdrawal rate and leave a higher portfolio final balance.

By far the simplest strategy is Bonds First and there is a strong logic behind the approach - leave the equities alone as long as possible to grow and use up the stable but slow growth bonds. However, it is virtually guaranteed that at some point the portfolio will be 100% equity. Other strategies such as Prime Harvesting may result in a significant bias towards equities but not to the extent of Bonds First.  However, it is far more complicated to execute.

McClung´s analysis of Prime harvesting from 1927 with a 60/40 equity/bond portfolio in drawdown shows that it is only during the late 1950s to early 1970s that bond holdings fall below 30% - but the possibility of this occurring may be too uncomfortable for some retirees. Although it is possible to set a floor to the bond holdings, selling equities when necessary to bring the bond holdings back to say 30%, this does reduce portfolio performance so conventional annual rebalancing may be preferable for the risk-averse.

I think that for many DIY investor retirees an important consideration has to be what happens if the day arrives that he is not capable of managing his financial affairs.  Entering retirement with a complex withdrawal strategy may be OK for a decade or so but at some point, a simplification of finances may be enforced.  This would make me reluctant to adopt a strategy such as Prime Harvesting.  Bonds First would be less complex and very effective but despite the analysis that shows a glide path to 100% equity produces superior results I am not sure I would be comfortable with this.


Post a Comment


Paul said…
Hi Max,
Great post and lots of useful info - many thanks.
I’m a year or so away from drawdown however, trying decide on a realistic, manageable and palatable investement / drawdown strategy is not at all easy! It seems that a UK investors SWR is much lower compared to the US. That coupled with current historically high market valuations (especially US), low bond yields and sustained inflation on its way it feels that sequencing risk ie a major crash in the short term just as retirement drawdown commences is a major consideration. In an attempt to address this posibility I have started looking at Tactical Asset Allocation strategies suitable for a retirement portfolio. From what I have found so far backtested results (US data) produce higher CAGR’s / Sharpe / Sortino much higher SWR’s and for most strategies much lower volatility and capital drawdowns than buy and hold strategies. The downside appears to be potential monthly adjustments to asset allocations depending on the strategy rules. Have you looked into these strategies. (Allocate Smartly has some useful info on TAA). If so it would be good to hear your thoughts.
Thanks again,
Max said…
Hi Paul, many thanks for your great comments. I haven´t looked at Tactical Asset Allocation - but I will, and will then comment. You´re right about UK SWR being lower than the US - we have to accept 3% to 3.5% is more realistic than the 4% which has worked well in the US. As you say US valuations are high, UK less so, but having said that the historic and Monte Carlo SWRs should still be valid independent of what part of the cycle you start drawdown. I believe that providing you enter drawdown with a realistic SWR the biggest problem is that you will end up not taking sufficient from your portfolio and when you do realise this you'll probably be too old to appreciate the money available! I think this is where variable drawdown should play a role - ensuring you maximise the drawdown whilst ensuring portfolio survival. I have been simulating various strategies and will be publishing a post in a few weeks. Kind regards Max