We run through the Monte Carlo simulation and how you can use it to understand how asset allocation and other variables in your portfolio will determine the longevity of your retirement portfolio. You can use these outcomes to improve your retirement outcomes.

You are able to find the full article here.

Listen on:

You can find the full transcript below:

Shani Jayamanne: Welcome to another episode of Investing Compass. Before we begin, a quick note that the information contained in this podcast is general in nature. It does not take into consideration your personal situation, circumstances, or needs.

Mark LaMonica: Okay, we have a little bit of a listener requested episode today, but before getting into this, and this is relevant to the episode, what do you think about gambling? Do you gamble? Because I’ve never seen you gamble.

Jayamanne: Okay, Aussies are really into gambling. Do scratchies count as gambling. I guess they would.

LaMonica: Yes, but that’s not gambling because you make me buy them for you. So I don’t think that that counts.

Jayamanne: I buy them sometimes. I get them a lot as gifts.

LaMonica: You find them when you find money on the street. That’s when you buy them. You found money on the street in Redfern. We know where that money came from, and then you go and buy them.

Jayamanne: I do, yeah. I gamble once every four years. Normally, I’ll put a little bet on…

LaMonica: On the presidential cycle.

Jayamanne: On the election.

LaMonica: Yeah, who’d you bet for this time?

Jayamanne: No, the Aussie election.

LaMonica: Oh the Aussie election.

Jayamanne: Yeah, so I’ll let you know. Normally, I bet on the one that I don’t want to win. So either way, I have something. So...

LaMonica: Well, there you go. Welcome to bad leadership that you don’t want, but at least you won five bucks.

Jayamanne: Yeah, exactly.

LaMonica: Okay, well, you’ve got a bet coming up then.

Jayamanne: I do.

LaMonica: Will you tell people you don’t have to now, because I know you need to think about this, check polls as the election’s called and it gets closer. Will you tell people who you bet on?

LaMonica: Or is that automatically saying...

Jayamanne: No, because who I support, so...

LaMonica: I know who you support, but I won’t tell anyone.

Jayamanne: It’s not hard to figure out.

LaMonica: Yeah, but anyway, this is related to this episode today. So one of our listeners wrote me an email and talked about, I guess we’ll say a little bit about his personal life, what he likes to do in his spare time. And apparently, that is running Monte Carlo simulations on retirement outcomes.

Jayamanne: Do you think this is a common pastime Mark?

LaMonica: I don’t.

Jayamanne: Okay, do you do this? This seems like something you would do.

LaMonica: No, I did this for this podcast. And I did this for this article I wrote, but no, it is not something I do, Shani. Thank you. So it’s not a common outcome or not a common pastime, but it could lead to a better retirement. And that’s a good trade-off, because then you can do whatever you want, which might be even when you’re retired, run Monte Carlo simulations.

Jayamanne: What could be better than that?

LaMonica: Exactly. Okay, let’s start with some basics. What is a Monte Carlo simulation?

Jayamanne: So a Monte Carlo simulation is named after the famous casino and estimates the probabilities of different outcomes. This is useful because the future is a result of complex interplay between a lot of factors.

LaMonica: And for retirement, the outcome we are estimating, the probability of is first and foremost, will you run out of money before you die or not? And then we can get into the probabilities of dying with a certain amount of money.

Jayamanne: And this means we need to explore each factor that indicates how much money you will end up with in retirement. And you can probably guess most of the factors. The first is obviously how much you have when you start retirement.

LaMonica: That does make sense, Shani. The next is how much money you take out of your portfolio. Now, in retirement planning land, this isn’t a dollar figure. It is a percentage that’s known as the withdrawal rate. And what we’re really talking about here is the initial withdrawal rate. So we’ll say 4% as an example, because that is the most famous one.

Jayamanne: And that’s used to calculate how much comes out initially. If you have $1 million and a 4% withdrawal rate, you take out $40,000 the first year. But then you increase that amount by inflation each year to maintain a consistent standard of living. That means the next number that matters is inflation.

LaMonica: And obviously, this is not a set rule. You can adjust how much you take out. It can be less than inflation or more than inflation. We’ll get into that. But this is the baseline and standard approach.

Jayamanne: And the next input that matters is the return that you earn on your portfolio.

LaMonica: Which of course also makes sense. And the final input that matters is how long your retirement lasts. And of course, how long your retirement lasts is one of the biggest challenges because planning for a 20 year retirement and a 50 year retirement are two very different things.

Jayamanne: What are you planning for, Mark?

LaMonica: Well, once again, I’m weeks from death. So I’m going to retire and I have a four week notice period. So I will literally die at my desk now.

Jayamanne: Bleak. All right. So that is everything that you get to input into a Monte Carlo simulation. As a summary, it’s how much you have, the withdrawal rate, the inflation rate, returns and how long your retirement lasts.

LaMonica: So as we said, the Monte Carlo simulation spits out a probability of running out of money given the different inputs. But it is worth considering what inputs are used for things like returns and inflation because those are of course unknowable.

Jayamanne: And there are a couple of different routes. So we can go with this. One of the most famous uses of the Monte Carlo simulation in the personal finance space was done by William Bengen to come up with a 4% rule. Now, in that case, he used historical returns for different asset classes. And his, of course, means you need to add another input, which is the asset allocation of this portfolio.

LaMonica: And you can use projected return data as well. To do this, you need both the average returns you expect in the future and the standard deviation of those returns that you expect in the future. So standard deviation simply measures the dispersion of returns around the average return. It is a measure of volatility. And this matters because as we’ve often talked about on this podcast, it isn’t just the average return, but also the order or sequence of returns. And that is called sequencing risk.

Jayamanne: And if you have a basic understanding of statistics, you know that the dispersion of those returns around the average return is based on a normal distribution. That follows the 68, 95, 99.7 rule, which means that 68% of the time, the returns will be within one standard deviation, up or down of the average return. 95% of the time, within two standard deviations of the average return and 99.7% chance of returns falling within three standard deviations.

LaMonica: Okay, stats class is over, Shani.

Jayamanne: It was my least favorite class in university.

LaMonica: Well, there you go. And I didn’t take it in university. But let’s get back to retirement class. Probably a class you did not take in university.

Jayamanne: No.

LaMonica: Yeah, me neither. But we’re back to retirement class. So we’re going to run our first simulation. And if you want to do this, you can do this on your own, like the listener did. So we’ll put a link in the episode notes, and you can go in and spend your weekend running Monte Carlo simulations. So for a baseline scenario, we’re using a 4% withdrawal rate for a 30 year retirement. We have a portfolio allocation of 35% to Aussie shares. Now, these are called Pacific shares in the simulator. So it does include New Zealand as well. 20% to U.S. shares, 15% to global ex-U.S. shares, 30% to U.S. bonds, just because there aren’t Aussie bonds in the simulator. And you put this all in, you press the calculate button, and you get a number. And the number is 78.37. So what does that mean, Shani?

Jayamanne: It all depends on how you feel about a 78.37% chance of not running out of money in a 30 year retirement.

LaMonica: Yeah, that would probably make you anxious.

Jayamanne: It would make me very anxious. But I’m going to change the inputs. And in the simulator, I’m going to put in 100% certainty of not running out of money. If you keep all of the inputs the same and change to a 12 year retirement, there is a 99.98% chance that you don’t run out of money.

LaMonica: Okay, that’s a short retirement.

Jayamanne: Yes, but it’ll be good.

LaMonica: Yeah, I mean, it could be really fun. You don’t have to take care of yourself. No gym, you take up smoking, drink as much as you want.

Jayamanne: But I do like the idea of a 30-year retirement. If you reduce your withdrawal rate from 4% to 1.85%, it gets you a 99.46% probability of not running out of money in a 30-year retirement.

LaMonica: Okay, but a 1.85% withdrawal rate. So basically, you can’t do anything.

Jayamanne: Yes, unless you have a lot of money.

LaMonica: Unless you have a lot of money.

Jayamanne: Yes. Well, one thing that we always talk about on this podcast is that investing involves exchanging volatility for higher returns. The more volatility you take, the higher the returns that you can expect to achieve.

LaMonica: Okay, that is true. So why don’t we turn up the volatility a little bit and we’ll allocate more of that portfolio to growth assets. So we’ll change the asset allocation to the following. So now we are putting 90% in shares. So 50% in Aussie shares, 20% U.S. shares, 20% in global ex-U.S. shares, and then we’ll put that remaining 10% in bonds.

Jayamanne: And unfortunately, this didn’t work out too well. The probability of not running out of money is now 68.59%. Perhaps it’s a bit of a surprising result. After all, shares have a higher historic return than bonds. And shares also have more volatility. A portfolio with more volatility will go down more in bad years. That is sequencing risk when those bad years occur early in retirement.

William Ton: I’m Will, producer of Investing Compass and here are this week’s must reads on Morningstar.com.au. There is a heightened anxiety about whatever is going on in Washington. There is a sense that we are at an inflection point between the era of globalization and one with more protectionism and economic nationalism. And there is a simple fact that negativity sells. Mark’s Unconventional Wisdom column explores the speculation in the financial media that we are now on a cusp of a lost decade. Shani’s Future Focus column runs through the biggest detractor of wealth for long-term investors. In the first of a two-part series, Shani lays down the foundation to show the twin reasons for why Aussies are feeling the cost of living pressures particularly in the last few years and why you should invest.

Shani also puts together a free calculator to provide investors with their personal inflation hurdle rate to work out what they need to maintain the purchasing power of their money. What can investors learn from the world’s second richest man or third or fourth depending on the week? In this week’s Bookworm, Joseph explores the distinction that Jeff Bezos draws between renting stocks and acting like a business owner. Younger generations are highly motivated to achieve financial independence, saving for retirement earlier than previous generations while also taking greater risks. With market noise and volatility at a high, Sim explores market bubbles and speculative trading in the case of Tesla and discusses what investors should do during periods of uncertainty. These articles and more they are now available in the show notes and let’s get back to Mark and Shani.

LaMonica: Okay, we’re going to go in the opposite direction we did before. You tried to get to as close to 100% certainty as possible to deal with your anxiety. Now we’re going to make your anxiety worse. We’re going to try to make this outcome as bad as possible. So you do have the opportunity in this Monte Carlo simulator to adjust the order of returns and it allows you to stress test different scenarios. For instance, you can select worst 10 years first. So remember sequencing risk? This is sequencing risk on steroids. And this means that you are terribly unlucky with the time that you actually retire. So in a 30 year retirement, which we’re using, if the 10 worst years of returns happen in the first 10 years, there is a 1.63% probability of not running out of money over 30 years. So that is not great. But that shows the impact of sequencing risk if you’re really, really unlucky.

Jayamanne: But we do have a secret weapon we can use and that is cash. Let’s go back to the original scenario. But this time we’re switching the 30% allocation from bonds to cash. The probability of not running out of money has increased from 78.37% to 80.24%.

LaMonica: So once again, this is that interplay between volatility and returns. And this time it works in our favor. Bonds have higher returns than cash. Yet cash does not have any volatility. So we’re starting to get somewhere, Shani.

Jayamanne: All right. So why don’t we lower the cash and increase the allocation to shares? The new allocation is 40% Aussie shares, 25% U.S. shares, 20% Global ex-U.S. and 15% cash.

LaMonica: Okay. So in this case, the probability of not running out of money slightly increases to 80.37%. But this is only part of the picture. As we mentioned at the beginning of the podcast, we can also look at return scenarios.

Jayamanne: We can see the different pathways for a portfolio by examining the baseline scenario. So as a reminder, that was 70% equities and 30% bonds, a $1 million account balance and a 4% initial withdrawal increased by inflation each year.

LaMonica: So that probability, if you remember, was 78.37%. In that same scenario, you have a 10% probability of ending up with more than $8.5 million in 30 years. The middle 50% of scenarios between the 25th percentile and 75th percentile shows an outcome of having between $248,000 and $4.53 million in your portfolio.

Jayamanne: All right. So let’s compare that to the 85% equities and 15% cash scenario using the same assumptions on withdrawals and portfolio size. At an 80.37% probability of not running out of money, you have a 10% probability of ending up with more than $13.1 million in 30 years. The middle 50% of scenarios between the 25th percentile and the 75th percentile shows an outcome of having between $349,000 and $6.6 million in your portfolio.

LaMonica: So that’s a significantly better outcome. And that previous scenario, remember that you have a slightly better chance of not running out of money in this better outcome than that baseline that we used. So the message seems pretty clear to me that more cash is better or some cash is better than more bonds.

Jayamanne: And the range of different levels of wealth at the end of the 30-year period shows the hand of fate on retirement. There’s a 10% chance that everything could go very right, which would result in more than $13 million in the second scenario. If everything goes really wrong, you have a 20% chance of running out of money.

LaMonica: Yeah. But there’s still more we can do by examining in the way the Monte Carlo simulation works. So the simulation assumes that you are selling equal parts of your portfolio to fund each withdrawal. So simplistically, this means if you had a portfolio of 50% shares and 50% cash, you would take half of your withdrawal from each bucket.

Jayamanne: And in the simulation world, this occurs if the shares went up 20% or down 20%. In the real world, you can selectively pick how to fund a withdrawal. In the scenario with 15% cash, there was an 80% probability of not running out of money. Yet if there is an 85% allocation in shares and those plunged, you couldn’t sell, but you do have cash.

LaMonica: The cash could fund your retirement spending for more than 36 months, even if inflation was high. This would give you time for the shares to recover. Since the end of World War II, there have been 13 bear markets. To get back to the market peak, the break-even point took an average of 21 months.

Jayamanne: So in the real world, dividends could fund at least part of your withdrawals. In the real world, you could adjust your spending in years where the market fell significantly and in years of high inflation.

LaMonica: So the point is, I think that this is valuable. This Monte Carlo simulation is valuable because it does show you how sequencing risk works, how different asset allocations impact your portfolio when you’re actually withdrawing money. But it also is not real world because you can actually adjust a lot of this. And so I think it’s good to run these, but the person that emailed me in was very worried that he could not come up with a realistic way to get to 100% probability of not running out of money. But that is based on the simulator. In real life, there are other things you can do. So how is that as a gambling exercise, Shani? Did you enjoy that more than…

Jayamanne: I feel like gambling is less calculative than that for most people.

LaMonica: Yeah, no. Mostly it’s a couple cocktails and a trip into the poker room. But anyway, thank you very much. We will link the article. We know there were a lot of probabilities and percentages and everything else, and that we’ll link an article that goes through all of this. But anytime you have a suggestion for an article, send it in to us. We would appreciate it.