Do Target Date Funds and the Bond Tent work for Retirement in Singapore?

One way of retirement planning is to:

1. Estimate the expenses needed in retirement (usually as a percentage of current expenses)
2. Assume that retirement lasts for 20 years
3. Make some assumptions about inflation and returns on assets
4. Work out the lump sum needed on the first day of retirement
5. Put your money in Target Date Funds to achieve the lump sum needed

This looks straightforward, right? The focus is on making sure that the lump sum is achieved, and not on how the money is allocated.

Target Date Funds

Target Date Funds are commonly used to get this retirement lump sum. These funds commit to grow to a certain value by a specific date, say by 2045. The worry for the fund is that market volatility close to the target date will reduce the fund’s value. So, the Target Date fund starts off with a high allocation to stocks, and gradually reduces this towards the target date, so the fund will be largely invested in bonds at the target date.

Illustration of the asset allocation in a 2045 Target Date Fund

What is the retiree supposed to do with the lump sum received from the target date fund? One option would be to leave it in the allocation of the fund, say, 90% in bonds and 10% in stocks. The downside to this is that, based on simulations, the portfolio allows only a 3.35% safe withdrawal rate before running out of money after 30 years.

Success Rates of a portfolio with 90% in bonds

Withdrawal Rate	20 years	25 years	30 years	35 years	40 years
3.5%	100%	100%	85%	31%	4%
3.35%	100%	99%	95%	52%	12%
3.0%	100%	100%	100%	91%	55%

Median values of a portfolio of 90% bonds (starting value $1m)

Withdrawal Rate	20 years	25 years	30 years	35 years	40 years
3.5%	$660,000	$468,000	$211,000	$0	$0
3.35%	$714,000	$544,000	$316,000	$14,000	$0
3.0%	$840,000	$722,000	$558,000	$334,000	$40,000

The high allocation to bonds earning a low rate of return (assumed to be 2.8%) means that the portfolio value will fall over time, leaving less in liquid reserves towards the end of the 30 years, precisely when they are needed for healthcare expenses and for bequests. In fact, the retiree would be better off putting the entire lump sum in an annuity. An annuity would not leave any liquid reserves for those end-of-life emergencies either, but will have a higher withdrawal rate, and does not run out after a 30 years.

What about portfolio allocation for FIRE?

An alternative is to put a larger amount into stocks, recommended by the FIRE (Financial Independence, Retire Early) movement. In theory, the higher allocation to stocks allows the portfolio returns to beat inflation, so it will not run out of money, even after 30 years. But as we show in Does the 4 percent rule work in Singapore, it does not work for us. But a portfolio with 50% in stocks can only support a 2.15% withdrawal rate over 30 years. Although the withdrawal rate is lower, the median value of the portfolio is much higher than the one with 90% in bonds. This means that liquidity for future emergencies can be met using this allocation.

Success rates and median values of a portfolio of 50% bonds / 50% stocks (starting value $1m)

Withdrawal Rate	20 years	25 years	30 years	35 years	40 years
2.15%	100%	98%	95%	89%	84%
Portfolio Value	$1.4m	$1.5m	$1.6m	$1.7m	$1.8m

The advantage of a FIRE-type asset allocation lies in the preservation and growth of the portfolio’s value. This is critical for early retirement, where the funds need to sustain expense for more than 30 or 40 years. Even for a retirement of 20 to 30 years, this helps in meeting end-of-life healthcare expenses, and family bequests. However, this comes at a lower sustainable safe withdrawal rate, which means a larger portfolio is needed.

What about the Bond Tent? Can it work?

While recognising the need to increase the portfolio allocation to safe assets like bonds just before retirement to avoid sequence of returns risk, some have argued that to sustain retirement, the allocation of the portfolio after retirement needs to be shifted back into stocks gradually. For example, Michael Kitces talks about this here. Big ERN also does so here and here. The changing retirement portfolio asset allocation, before and after retirement, is a Bond Tent.

Illustration of the asset allocation in a Bond Tent (assuming retirement in 2045)

The Bond Tent has worked in the US in the past. This is due to the strong performance of the S&P 500 stock index historically. The question is, does the Bond Tent work elsewhere?

Testing the Bond Tent using simulations

As before, we fired up our trusty retirement simulator to see if the Bond Tent works for a retiree in Singapore re-allocating the target date fund upon retirement. We assume the same parameters as before:

Assumptions used for simulations

	Annual Returns	Standard Deviation of Annual Returns	Annual Fees
Stocks	5.07%	26.32%	1.00%
Dividends	2.80%	1.09%	0.00%
Bonds	2.80%	0.00%	0.00%
Inflation	2.00%	0.00%	NA

We also assume that the portfolio starts with an allocation of 90% in bonds and 10% in stocks, and this allocation changes by 5% each year until it reaches an allocation of 50% in bonds and 50% in stocks in 8 years. What do you think will be the outcome? Even without doing the simulations, we know from above that a 50:50 stock-bond allocation only succeeds with a 2.15% safe withdrawal rate for 30 years, so it is very unlikely that this Bond Tent portfolio asset allocation can do much better than that.

And that is the case We now have a slightly higher safe withdrawal rate of 2.4%, but at the expense of a smaller median portfolio value. In fact, these withdrawal rate and median portfolio value outcomes lie in between the results for a 90% bond allocation in a Target Date Fund, and the 50% bond allocation in a FIRE-type portfolio.

Success rates and median values of a Bond Tent allocation portfolio (with a starting value of $1m and 90% in bonds, and ending with 50% in bonds)

Withdrawal Rate	20 years	25 years	30 years	35 years	40 years
2.4%	100%	99%	95%	89%	82%
Portfolio Value	$1.3m	$1.3m	$1.3m	$1.3m	$1.3m

The Bottom Line

So the bottom line is that the Bond Tent does not work for retirement portfolio asset allocation in Singapore, and probably not in most other countries as well. It has worked in the US, thanks to the strong performance of the S&P 500 stocks historically. However, that may not be so in the future, especially in the low interest rate environment where bond returns are low.

Quick takeaways:

1. Target Date Funds allocate more to safe assets like bonds as they approach the target date. This is to safeguard against sequence of returns risk
2. However, a high allocation to bonds is not feasible for retirement, as it gives a lower payout than an annuity. It also exhausts the funds after 30 years
3. A higher allocation to stocks of 50% does not exhaust the portfolio, but has a lower safe withdrawal rate
4. A Bond Tent, which gradually shifts away from bonds towards stocks is a compromise. It preserves the value of the retirement portfolio and has a slightly better withdrawal rate
5. Neither Target Date Funds nor a Bond Tent work well for retirement. It is better to combine an annuity with a Yield Shield portfolio for retirement. This is described here and here

Also take a look at our previous posts:

• Does the Yield Shield protect retirement finances?
• The Yield Shield, 4 Percent Rule and CPF Life – A Perfect Retirement Combo?