The Disadvantage of Using Montecarlo Simulation in Investment Analysis.

Monte Carlo simulations, while powerful for modeling complex financial scenarios, can present certain risks in investment. One significant danger is that the assumptions underlying the simulations may be overly optimistic or not accurate, leading to unrealistic expectations and potentially jeopardizing the investor's financial goals. Additionally, the limited historical data used to generate these simulations can lead to a false sense of security, particularly during economic downturns or major market shifts. Moreover, retail investors may lack the expertise to interpret the results correctly, making it difficult to make informed decisions based on the simulation outcomes.

A few key factors of disadvantages using Montecarlo Simulation

A) Historical Data & Normal Distribution

Montecarlo simulations, while used mostly in financial planning and valuation, have certain potential disadvantages when used in asset management, pension funds, and other investments, particularly for retail investors. One key concern, as mentioned before, is the reliance on historical data and the assumption of normal distribution, which may not accurately predict future market behavior. Also, these simulations might foster a false sense of security due to their mathematical nature, leading investors to underestimate the inherent risks in their investment strategies. Lastly, retail investors may lack the technical expertise to correctly interpret the results of Montecarlo simulations, potentially leading to unsound investment decisions.

B) Over-Optimization

Another major concern with using Montecarlo simulations in asset management and investment planning is the potential for over-optimization. This happens when investors become overly reliant on the simulation results, tweaking their models constantly to optimize outcomes—a phenomenon known as "curve-fitting." As a result, investors may design strategies that work well in the simulated environment but underperform in real-world conditions. This over-optimization can be particularly detrimental for retail investors who may have limited resources and experience in financial management, leading to high costs or even financial ruin in extreme cases.

C) Overfitting

Another significant concern is the possibility of overfitting, which occurs when a model is tailored too closely to past data and may not perform well in future scenarios. This problem is particularly relevant to retail investors, who may not have access to advanced models or expert guidance in adjusting their simulation parameters. For instance, if a Montecarlo simulation excessively relies on a specific market trend that becomes obsolete, retail investors who follow this model may suffer financial losses. Therefore, it is crucial for retail investors to recognize the limitations of Montecarlo simulations in the context of investment decision-making and avoid placing undue weight on the results.

D) Inability to detect Black Swan events

Furthermore, even when Montecarlo simulations are utilized with the utmost care and expertise, their inability to account for unexpected events or "black swans*" poses a considerable risk for those who rely heavily on their projections. In the realm of asset management, pension funds, and other investments, unforeseen crises or major regulatory changes can have a severe impact on markets and investment strategies. For retail investors, these potential vulnerabilities in Montecarlo simulations may lead to insufficient diversification or risky portfolios that fail to protect their long-term financial interests. This emphasizes the importance of pairing such simulations with comprehensive investment analyses and maintaining agility in the face of ever-changing market conditions.

Here is a great example from advisors Massimo Young, CFA** and Wade Pfau Ph.D., CFA, RICP?*** extracted from VettaFi (View their entire articule and the methodology/assumptions used for this example in this link).

Jane is 65 and just retired. She has $1 million in an IRA, and her current allocation is 60% U.S. large-cap stocks and 40% U.S. investment-grade corporate bonds. She works with her advisor to arrive at a prudent portfolio and withdrawal strategy to make sure she does not run out of money by age 95.

The advisor uses a planning tool to run 1,000 Monte Carlo simulations, evaluating different withdrawal rates as well as different portfolios for Jane. Standard practice at the advisor’s firm is to use 80% probability of success as the cutoff – anything less than that is not considered “safe enough.” The results show that if Jane keeps her current portfolio, she can spend $41,200 next year, and 2% more each year after that to adjust for inflation, with an 80% probability of success to age 95.

This result is roughly consistent with the “4% rule****” used by many advisors, which Jane may have heard of. She goes home confident in her plan and reassured that it is “safe.”

Is Jane right to feel secure in this answer?

“Monte Carlo analysis” and “probability of success” sound like highly technical and scientific terms. But what’s behind the math Probability of success is the percentage of Monte Carlo simulations where the client a) spends a given amount each year and b) still has money left over at a given age. An 80% probability of success means that in 80% of simulations, Jane still has money at age 95; in 20% of simulations, she runs out of money before age 95.

Probability of success, therefore, depends on how the Monte Carlo analysis is set up. Monte Carlo is a technique for generating a set of future scenarios (“simulations”). In the case of retirement income, the analysis generates, say, 1,000 simulations of a portfolio, given a withdrawal strategy. But it is not?all possible future values,?or even a completely “random” selection of future values. The set of future values evaluated is determined entirely by the assumptions fed into the analysis.

In each year of a simulation, the Monte Carlo analysis will assign a return to the portfolio: “the return in year 1 will be X%.” This X% is selected randomly from a?pre-determined,?specific distribution of returns?of the portfolio. But the distribution itself is not random. The distribution is determined by the assumed returns, volatilities, and correlations among asset classes (CMAs). For example, if the average return in the distribution of portfolio returns is 7%, there will be a much higher chance that 7% (or something close to it) is selected as the X% rather than, say, 60% or -15%.

Since these are all forecasts, Jane might reasonably ask her advisor: What if you’re wrong?

Using 2022 Horizon CMA Survey data, the results suggest that, given her allocation, the “safe” spending level could be anywhere between $33,000 and over $51,000, just depending on which stock return assumption is used (and keeping everything else the same).

Put differently, if her advisor plays it safe and uses the lowest return prediction, Jane could be underspending by 56% (if the highest return forecast turns out to be the right one). Alternatively, if her advisor aligns with the most bullish forecast, she may be overspending by 36% and have a good chance of running out of money early (if the lowest return forecast turns out to be right).

Another way to look at this is to evaluate the 4% rule in the context of different CMAs. What is the probability of success of the 4% rule using the different return forecasts?

Using the most bullish estimate, Jane’s probability of success using the 4% rule would be over 95%. Using the most pessimistic, it would be just 62%.

But smaller differences matter too. If Jane’s advisor used 7.11% (25th percentile) instead of 8.52% (75th percentile) for stock returns, Jane’s probability of success would fall from 87% to 78%. Put differently, her estimated probability of failure – running out of money completely by age 95 – would increase from 13% to 22%. That’s a 70% higher chance of failure, just based on a return difference of 1.4 percentage points. And that’s just for one input into the Monte Carlo analysis.

All in all, Montecarlo simulations have become increasingly popular in the financial industry, particularly for asset management, pension funds, and other investments. These simulations provide a method for evaluating the potential risks and rewards associated with various financial decisions by generating multiple probable outcomes. However, despite their increasing use, Montecarlo simulations also have inherent risks, especially for retail investors who may lack the knowledge and expertise to fully understand the implications of the results given or its reliability.

* A Black Swan is an extremely negative event or occurrence that is impossibly difficult to predict. In other words, black swan events are events that are unexpected and unknowable.

** Massimo Young, CFA leads investment solutions and technology for Insight Investment’s Individual Retirement Solutions group. Note: The views expressed in this article are those of the authors are do not necessarily reflect the views of Insight Investment.

*** Wade D. Pfau, Ph.D., CFA, RICP?, is the program director of the Retirement Income Certified Professional? designation and a Professor of Retirement Income at The American College of Financial Services in King of Prussia, PA, as well as a co-director of the college’s Center for Retirement Income. As well, he is a Principal and Director for McLean Asset Management and RISA, LLC. He also serves as a Research Fellow with the Alliance for Lifetime Income and Retirement Income Institute.

**** The 4% Rule is a Rule of thumb and states that you should be able to comfortably live off of 4% of your?money in investments?in your first year of retirement, then slightly increase or decrease that amount to account for inflation each subsequent year. For example, in the first year of retirement, you can withdraw up to 4% of your portfolio's value. If you have $1 million saved for retirement, you could spend $40,000 in the first year of retirement following the 4% rule.