## Does the 4 percent rule work in Singapore?

In our previous blogpost * The 4 Percent Rule for Financial Independence*, we discussed the “safe” rule which is popular with the FIRE (Financial Independence, Retire Early) movement. The 4 percent rule is a rule of thumb on how to create an annuity from a portfolio of assets, and has become the received wisdom on how to live on the stream of income it produces. A variant of the rule is that once you saved 25 times your annual expenses, you would be able to retire.

However, Finke, Wade and Blanchett (and also here) argued that the fact the 4 percent rule worked using historical US data was an anomaly, in a period where both stock and bond returns were higher than they are today in the US. If future stock and bond returns are lower, as they would be in the low interest rate environment of the last decade, then retirees should instead be using a much lower 2.5% initial withdrawal rate to be assured of retirement sustainability. Work done by “Big Ern” here also support this conclusion.

If true, this is bad news for retirees who hoped that the 4 percent rule would give them a comfortable standard of living in retirement. Now, they can only look forward to a more frugal existence, or run the risk of outliving their funds.

#### Testing the 4 percent rule Singapore

So far, most of the research on safe withdrawal rules has been focused on the US, and ignored other features about retirement – such as the use annuities (like CPF Life), life expectancy (the chance that a retiree would actually survived for 30 years in retirement, from age 65 to 95). Abstracting from these features for now, we want to know does the 4 percent rule works in Singapore for someone retiring at the age of 65 today and expecting another 25 to 30 years in retirement. Would retirement be a comfortable experience, or a terrifying one, wracked by fears of outliving his/her money, or being forced into a frugal existence?

Since we do not have data for Singapore stock and bond returns since 1928, there is no way to do a similar historical exercise on the feasibility of the 4 percent rule in Singapore. However, a Monte Carlo simulation based on the projected returns on stocks and bonds over the past 20 years for which data has been available can be done.

Singapore stock returns have been quite volatile over the past 20 years. Using the Straits Times Index (STI) returns over 20 years, we assume that the average annual return on stocks is ** 5.07%**, with a standard deviation of

**. From the past 10 years of the dividend yield on the STI Exchange Traded Fund (ETF), we estimate the average dividend yield per year to be**

*26.32%***with a standard deviation of**

*2.80%***. For bonds, we assume the yield on bonds available for retail investors would be similar to the yields on the Singapore Savings Bonds, or very generously around**

*1.09%***per year, which has been the highest projected for a ten year holding period in the past few years. The correlation between all these returns is assumed to be zero.**

*2.80%*While we can quibble whether these assumptions are 100% accurate, basing them on the actual returns and volatility seen over the past 10 to 20 years assures us that they are reasonable estimates of the investment outlook in the next 30 years. Here is a summary of the initial assumptions, returns, standard deviations and fees used:

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 |

#### Let’s run some simulations!

For our first set of simulations of annual stock returns, dividends and bond interest, we assume an asset allocation of 50% in stocks and 50% in bonds, re-balanced annually after the withdrawals from the portfolio are made at the start of the year. We further assume that the withdrawal amount rises by the rate of inflation, 2%, every year. Failure in the withdrawal plan occurs when the value of the portfolio reaches zero before the end of the retirement horizon (at 20 years, 25 years, 30 years, 35 years, and 40 years) in the simulation.

10,000 simulations are run in total, and the percentage of success across all of these simulations is tallied in the table following:

Withdrawal Rate | 20 years | 25 years | 30 years | 35 years | 40 years |
---|---|---|---|---|---|

5.0% | 65.46% | 44.84% | 30.91% | 22.52% | 17.12% |

4.5% | 76.22% | 66.62% | 41.87% | 31.48% | 24.81% |

4.0% | 85.51% | 69.17% | 54.40% | 43.20% | 34.73% |

3.5% | 92.42% | 80.49% | 67.78% | 56.62% | 47.47% |

3.0% | 96.80% | 89.48% | 79.67% | 70.75% | 61.89% |

2.5% | 98.95% | 95.70% | 91.41% | 82.31% | 75.84% |

2.0% | 99.79% | 98.71% | 96.27% | 96.22% | 87.46% |

Oh dear! As can be seen above, if we define safety as a success rate of 95% or more, then only withdrawal rates of 2.5% are safe for a 25 year retirement horizon, and a 2.0% withdrawal rate is safe for a 30 year retirement horizon. Even over a 20-year retirement horizon (which is short given that life expectancy is 85 years, meaning that half of the retirees would still be alive at age 85), only a 3.0% withdrawal rate, or a little higher, would be considered safe. Also, for those hoping to achieve financial independence and early retirement in Singapore, the answer seems to be that the safe withdrawal rate is around 1%! This is definitely bad news for those who believe that the 4 percent rule will work in Singapore.

For a portfolio half in stocks and half in bonds, only a 2% withdrawal rate is safe for a 30 year retirement horizon

#### More in stocks? Or less?

Would the 4 percent rule be rescued by allocating more of the portfolio to stocks, as stocks have higher returns to help grow the portfolio over time? The simulations with a 60% allocation to equity are shown below:

Withdrawal Rate | 20 years | 25 years | 30 years | 35 years | 40 years |
---|---|---|---|---|---|

5.0% | 62.77% | 45.46% | 33.64% | 25.94% | 21.19% |

4.5% | 72.20% | 55.24% | 42.93% | 33.89% | 28.14% |

4.0% | 80.66% | 65.89% | 53.28% | 43.81% | 36.93% |

3.5% | 88.00% | 75.72% | 64.74% | 54.80% | 47.42% |

3.0% | 93.75% | 84.48% | 75.26% | 66.95% | 59.35% |

2.5% | 97.17% | 92.02% | 84.74% | 77.81% | 71.90% |

2.0% | 99.11% | 96.70% | 92.59 | 87.62% | 82.67% |

These outcomes are even worse than before! While stocks have higher returns, the higher volatility seen over the past 20 years e.g. Asian Crisis in 1997/8, SARS in 2003, Global Financial Crisis in 2008, mean that a higher allocation to stocks would actually be less safe in retirement.

What about lowering the asset allocation to stocks? The results of the simulations with a 40% allocation to stocks are shown following:

Withdrawal Rate | 20 years | 25 years | 30 years | 35 years | 40 years |
---|---|---|---|---|---|

5.0% | 68.20% | 42.60% | 25.80% | 16.70% | 11.14% |

4.5% | 80.73% | 57.39% | 38.85% | 26.74% | 19.32% |

4.0% | 90.51% | 72.81% | 54.50% | 40.41% | 30.53% |

3.5% | 96.29% | 85.25% | 71.00% | 67.22% | 45.91% |

3.0% | 98.97% | 95.00% | 84.38% | 73.98% | 63.97% |

2.5% | 99.82% | 98.28% | 94.11% | 87.32% | 80.16% |

2.0% | 99.99% | 99.72% | 98.69% | 96.20% | 92.24% |

Now, contrary to the 4 percent rule findings in the US which require an allocation of at least 50% to stocks, we start to see that an initial 3.0% withdrawal rate may just be borderline feasible in Singapore over a 25-year retirement horizon, but for a 30-year horizon, a 2.5% withdrawal rate is still required.

So far, everything looks pretty bleak for retirees. But perhaps we have estimated the fees for investing in stocks too high at 1%. Maybe 0.25% in fees for a buy and hold investor would be sufficient. Maybe, we don’t need to index the withdrawals for inflation, since retirees tend to spend less and less as they get older. Can the 4% withdrawal rule be saved? The simulation results assuming 0% inflation adjustment and 0.25% stock investment fees are shown below. The results are still not promising, and the 4% rule can only apply for a retirement horizon of 20 years, which is too short given our longevity.

Withdrawal Rate | 20 years | 25 years | 30 years | 35 years | 40 years |
---|---|---|---|---|---|

4.0% | 96.04% | 89.62% | 82.38% | 76.02% | 71.27% |

#### Taking a closer look at what is happening

At this point, let’s take a detour to understand why the 4 percent rule is working so poorly in simulations for Singapore. The chart below shows the evolution of the value of the portfolio over time as for the various simulations, based on the withdrawals being made and the returns on the portfolio, split by percentiles.

- At the top, we see that the luckiest 30% of the simulation runs (the blue and orange zones) have an increase in portfolio value over time, so they essentially have no issues with funding retirement, and in fact can increase their spending above the 4%-of-portfolio-value initial amount.

- In the middle, we see the next 30% of the simulation runs (the grey, yellow and blue zones) also have no problems in making the money last for 35 years or more. The asset portfolio values increase initially, and may then dip down in the future when the withdrawals are larger, but there is little risk of outliving the funds.

- The bottom few zones are the simulation runs that have trouble making the money in the portfolio last for 25 years to 30 years. In particular, the worst case, which is the very bottom of the dark grey zone, says it all. In the first year of retirement, the portfolio suffers a loss of 40% of its value. This loss, together with the withdrawals from the portfolio, leaves a hole so large that no amount of future returns can make up for, and the portfolio runs out of money in the 10
^{th}year. Looking at the top of the dark grey zone, which is the bottom 10% of the simulation runs, we see that if it were possible to avoid making a loss of more than 15% of the portfolio value in the first couple of years of retirement, then the retiree has a good chance of making it past year 20 of retirement at a 4% withdrawal rate adjusted for inflation, which is the absolute minimum requirement.

This interplay of:

- withdrawals from the portfolio
- losses in the first couple of years of retirement, and
- no other sources of income

resulting in disastrous retirement outcomes is known as sequence of returns risk. Finding a solution to this risk is the key to designing a worry-free retirement, and we shall return to this in subsequent posts. There are no perfect solutions to sequence of returns risk, as retirement is defined as the point when the person has no other sources of income, needs to live off not just the gains and interest, but also the capital or principal of the asset portfolio, and when his asset portfolio is at it’s peak, and hence most vulnerable to a market downturn.

#### What can be done for retirees in Singapore?

So retirees in Singapore are caught in a difficult bind. Stocks are too volatile to ensure a safe retirement, while the interest rates on Singapore Savings Bonds are too low. But what about the CPF? If we did not withdraw our CPF, but instead left what we can in CPF, earning 4% interest in the Retirement Account (RA) and Special Account (SA), and use this as the “bond” allocation, can this make a difference? Does the 4 percent rule now work in Singapore using CPF?

The answer is yes! The results of the simulations with different allocations to CPF RA and SA are shown below. They effectively resuscitate the 4% rule when a majority of the assets are left in CPF. Keeping the funds in the CPF RA or SA helps the retiree to avoid the sequence of returns risk, as there is no downside to the RA and SA monies, preserving their value regardless of market conditions. At the same time, the CPF RA and SA accounts pays enough interest to sustain the retiree without dipping too much into the capital or principal sum.

Withdrawal Rate | 20 years | 25 years | 30 years | 35 years | 40 years |
---|---|---|---|---|---|

4.0%50% in CPF | 91.19%% | 79.15% | 67.68% | 57.66% | 49.66% |

4.0%70% in CPF | 98.68% | 90.61% | 75.83% | 60.57% | 47.41% |

4.0%100% in CPF | 100.00% | 100.00% | 100.00% | 0.00% | 0.00% |

3.5%100% in CPF | 100.00% | 100.00% | 100.00% | 100.00% | 100.00% |

In fact, when 100% of the assets are left in the CPF RA or SA, there is no chance of failure when a 4% withdrawal rule adjusted for inflation over a 30-year retirement horizon. And if we lower the initial withdrawal rate to 3.5%, we effectively would not run of money over 40 years! However, bear in mind that this only happens when we can get the 4% rate of interest in the SA or RA, and there are limits to how much we can top up these accounts. The Ordinary Account, which pays 2.5%, would not help, as it is even lower than the Singapore Savings Bonds considered earlier.

When 100% of assets are left in the CPF SA, the 4% rule works for a 30-year retirement

However, under a 4% withdrawal rate with 100% of the funds in the CPF RA or SA, the portfolio runs out between the 30^{th} year and the 35^{th} year, and so there can be little no bequest left to the retiree’s heirs should he/she go down this route.

#### What other ways do we have to retire securely?

Does this mean that the future for retirees in Singapore is bleak, trapped between high costs of living, low interest rates and volatile stock returns? We have not gone further into other approaches like:

- Withdrawing from an asset portfolio in combination with other financial products, such as annuities (e.g. CPF Life)
- The likelihood of survival until the age of 95, and
- The need to adjust the withdrawal amounts to inflation

All these other methods can help make the withdrawing from an asset portfolio more sustainable, and will be covered in later posts.

Do not rely solely on your savings

or asset portfolio for retirement!

However, the clear message here is: **Do not rely solely on your savings for retirement!** Unless of course you are a multimillionaire for whom a 2% withdrawal rate is more than sufficient for the lifestyle you want to live.

#### Reference

Finke, Michael, Wade D. Pfau, and David M. Blanchett. 2013. “The 4 Percent Rule Is Not Safe in a Low-Yield World.”* Journal of Financial Planning* 26 (6): 46–55

## One thought on “Does the 4 percent rule work in Singapore?”