MonteCarlo calculations

Questions:

I have a dilemma. Hopefully you can help. My investment adviser recently provided me with a retirement plan using another product.

Income, future assets etc. were comparable to what I calculate using RetireWare.

However, for the same scenario, RetireWare is indicating a 100% probability of success. The competing software shows a 14% failure. The adviser recommends up to 10% failure as acceptable.

I assume that both programs should yield generally comparable Monte Carlo results?

Answer:

Please note that there is a setting in the Options page that counts small shortfalls as a success, so by setting the threshold to $0, the probability may go down.

Also note that Naviplan probably uses a different approach for their simulations. For example, they may use only one expected return and volatility, or maybe only fixed income and equities for asset classes. In any case, with such calculations resting heavily on assumptions and methodologies, one may view a 14% failure as equivalent to a 10% failure. Also, since this is the product of an advisor, there is an incentive to err on the conservative side since your success is at stake, so he may be extra careful to recommend higher spending.

One way to assess your plan is to look at the various results produced by the software. You have the odds of success, a deterministic projection based on your selected expected returns (or RetireWare's standard, which is fairly conservative). You also have a projection assuming you earn poor investment returns in the future. Then there is the risk analysis that "stress tests" the plan against the main risks under various economic conditions.

Our idea in making decisions when facing an uncertain future is to take a "holistic" approach. If you see that most indicators are favourable, then you probably are OK.

One last thought. If you find you experience lower than expected returns or higher expenses after a year or two, you are able to correct the course by monitoring your plan and adapting your spending going forward to set you back on solid footing to meet your expenses throughout retirement.

By: RetireWare / Equisoft On 4:57 PM

Continue Reading

A few product questions

Question:

I have a few questions about some inputs that I cannot resolve on my own. I am a new subscriber but I think I have everything figured out now except the below.

Is there a way to model a spousal loan? for example, if I have lent my spouse $1 Million at 1% until death there is a $10,000 deductible interest expense for one and interest income for the other.

I don't quite understand the "dividend yield" and "increase in earnings" fields in the economic forecast. If I input all the nominal equity asset class returns as 7% for example, what happens when I manipulate the other two variables? it seems to effect the Monte Carlo simulation significantly so I want to be more certain as to their meaning.

For the Monte Carlo simulation, how conservative are the input volatility? I am trying to get more comfortable on the simulation and how I manipulate the variables to get an 85% chance of success or a 35% chance (or everything in between).

Answer:

To model the loan, you could add an 'Other Income; of $10,000 per year for you, and a $10,000 interest expense on her budget.

The "dividend yield" adds a constant dividend on the share of assets invested in equities (based on the profile selected in 'Asset Mix for Projections' on the 'Options' page. So if you select 2% and have $100,000 in equities, it will add $2,000 in the following year of dividend income taxed assuming they are eligible dividends. In future years, it will be based on the market value of equities, so as they grow, so does the dividends.

The "increase in earnings" field in the economic forecast is an assumption for annual wage increase. So if you are not retired and earn $100,000 per year with a 3% annual increase in earnings, the program will calculate earnings of $103,000 next year and increase it by 3% each year up to retirement. This is the only purpose of this assumption. In turn, your annual earnings are used to estimate the CPP.

The increase in earnings and dividend rate are not subject to volatility in the simulation. In other words they are assumed to be constant.

The default volatility for each asset class are based on historical experience adjusted to recent trends. In a nutshell, cash and fixed income are assumed to be less volatile than historically, and equities are in line with their historical volatility (based on the last 20 years), with International equities experiencing the most volatility. These are revisited annually.

By: RetireWare / Equisoft On 4:55 PM

Continue Reading

MonteCarlo Calculation

Question:

I have a dilemma. Hopefully you can help. My investment advisor recently provided me with a retirement plan using Naviplan. Income, future assets etc were comparable to what I calculate using RetireWare.

However, for same scenario Retireware is indicating 100% prob of success. He is indicating 14% failures. - He targets 10% max as acceptable.

I am not sure what he is using for Retirement Goal Tolerance - rest of the variables look comparable.
 Based on what it has taken to get to this point, I don't have a lot of confidence in his results and I would not bet my life on my results.

I assume that both Naviplan and RetireWare should yield generally comparable Monte Carlo results?

Answer:

Please note that there is a setting in the Options page that counts small shortfalls as a success, so by setting the threshold to $0, the probability may go down.

Also note that Naviplan probably uses a different approach for their simulations. For example, they may use only one expected return and volatility, or maybe only fixed income and equities for the asset classes. In any case, with such calculations resting heavily on assumptions and methodologies, one may view a 14% failure as equivalent to a 10% failure. Also, an advisor will tend to err on the conservative side since your success is at stake, so he may be extra careful to recommend higher spending.

One way to assess your plan is to look at the various results produced by the software. You have the odds of success, a deterministic projection based on your selected expected returns (or RetireWare's standard, which is fairly conservative). You also have a projection assuming you earn poor investment returns in the future. Then there is the risk analysis that "stress tests" the plan against the main risks under various economic conditions.

Our idea in making decisions when facing an uncertain future is to take a "holistic" approach. If you see that most indicators are favorable, then you probably are OK.

One last thought. If you find you experience lower than expected returns or higher expenses after a year or two, you are able to correct the course by monitoring your plan and adapting your
spending going forward to set you back on solid footing to meet your expenses throughout retirement.

By: RetireWare / Equisoft On 8:12 PM

Continue Reading

Monte Carlo Details

Question:

How are the economic scenarios used in the Monte Carlo simulations generated? Are they generated within the software, or does the software contain a set of hundreds of scenarios that are used when necessary?

I am trying to understand the stochastic projection component of this software better.

Any more information on this aspect of the software would be appreciated. Thanks.

Answer:

The random scenarios are generated assuming rates of returns of each asset class follow a lognormal distribution. Expected returns and standard deviations are based on historical information adjusted for current trends.

Random numbers are generated using Microsoft's random number generator. Each time the program runs a set of simulations the returns are generated based on the random number generator.

There is a blog post about Monte Carlo simulation:

http://retireware.blogspot.ca/2012/05/monte-carlo-simulation-is-mathematical.html

You will also find some information in the RetireWare help files:

https://secure.retireware.com/web/mc_help.aspx?language=en-CA&node=108&id=res_MChelp2.htm

On a related matter, this is a blog post about risk analysis and our approach:

http://retireware.blogspot.ca/2013/12/quantifying-risk.html

By: RetireWare / Equisoft On 10:42 AM

Continue Reading

Results

Question:

I'm struggling to interpret the results.  On one hand it looks like we our plan is in good shape.  But why is our probability of success so low?

Answer:

It is possible to have a situation in the simulation where bad investment returns cause shortfalls before you dispose of an asset such as the personal residence.

If you assume disposition slightly earlier, then these funds will be available as a source of retirement income.

In real life, if you became broke and owning a home debt-free, you would sell to access your equity.

By: RetireWare / Equisoft On 2:36 PM

Continue Reading

Monte Carlo Settings

Question:

In the Monte Carlo settings, I checked the box to "Use Rates of Return as specified in the Economic Outlook". Why are the rates that shown in the Monte Carlo settings different from the economic forecast and the fields disabled?

Answer:

When you select to use the economic basis (either the standard forecast or custom rates), these fields become disabled, as you need not make any selections in that location.

By: RetireWare / Equisoft On 2:35 PM

Continue Reading

Odds of success

Question:

When I run a Monte Carlo simulation, I get 92% success odds, but the assets at retirement show a shortfall.

When I look at the charts, it looks like there is/may be a slight shortfall near 85-90…

Why does this happen?

Answer:

Each capital market simulation looks at whether you have enough funds and income to pay your expenses for the rest of your life.

Suppose you have enough each year of retirement except that during the very last year at age 90 you are short by $10,000. I would consider this a success even though you are technically short by a small amount. After all, your selection of life expectancy is a best guess, or better, a safe guess.

There is an option in the Monte Carlo settings that allows you to count cases with small shortfalls as a success. You can modify the setting to only consider cases that have no shortfall, or select the amount you think is reasonable for your purposes.

If this does not explain your issue, I can have a look at your calculation if you give me permission to review your file.

By: RetireWare / Equisoft On 9:50 PM

Continue Reading

Monte Carlo Simulation and Withdrawal Amounts in Retirement

Question:

I have done extensive research on the Internet regarding using Monte Carlo simulators for retirement planning. From what I have read, it is more important to withdraw a fixed percentage (i.e 4%) of your total portfolio each year rather than a dollar amount. For example a $1 million porfolio you would withdraw 4% or $40,000 per year and let the rest grow.

My question is does your Monte Carlo simulation allow you to enter withdrawls amounts as a percentage of the portfolio as opposed to just a dollar amount. Some programs will event tell you how much you can withdraw at a given level of probability.

In other words, you have a 95% probability of not running out of money for xxx years if you withdraw xxx amount per year. How does your program deal with these issues.

Answer:

With RetireWare, you build a plan by choosing a retirement income objective, and the income need is met using Government and company pensions, other income, and investment assets, each with their withdrawal rules and tax treatment. You run your plan on a deterministic basis (that is, assuming no variations in expected rates of return).

Withdrawals from invested assets depend on what you need to close the gap between the retirement income objective and income payable each particular year. This amount varies as some are indexed, others come in payment only from a certain age and some cease being paid at a particular point in the future (for example, if you work on a part-time basis for a few years after retirement).

The retirement plan that you create with all this level of detail is "stressed-test" using a Monte Carlo simulation, where the expected rate of return for each asset class varies randomly in accordance with its volatility. Monte Carlo is integrated with the actual plan and is not merely a flat withdrawal percentage each year. We have seen articles showing a safe withdrawal rate using Monte Carlo simulation as an analysis tool, but they usually used a simplified model. Our model takes into account various types of income and various types of investment accounts (non-registered, RRSP, locked-in assets, etc., each with their own withdrawal constraints).

You can try different retirement income objectives and see the probability of success for each. Then you can decide the income level you feel most comfortable with.

By: RetireWare / Equisoft On 10:26 AM

Continue Reading

What is Monte Carlo Simulation?

Monte Carlo simulation is a mathematical method used to estimate the most likely outcome and the odds that certain events will occur, given a mathematical model of the problem.

Like the roulette wheels associated with the casinos of Monte Carlo, Monte Carlo simulation reproduces random outcomes by generating random numbers to obtain results. Unlike a roulette wheel, the Monte Carlo method uses random numbers to measure and quantify uncertainty and chance events.

The program generates random rates of return and develops a large number of potential future outcomes in the capital markets, assuming that past averages and standard deviations will hold in the future.

Expected returns from equity asset classes (such as Canadian, US or international equities) are typically higher than returns from low risk or risk-free investments (e.g., cash, GICs and fixed income). But higher equity returns also have greater risk, that is a greater range of outcomes, from complete loss of capital to appreciation many times over the initial purchase price. And they also experience greater volatility.

Probability of success

The probability (or chance) of success is the number of times that your plan is successful, i.e. the number of illustrations in which you have enough money to sustain your lifestyle until the end of your life, divided by the number of all simulations. For example, if the portfolio does not run out of money 800 times out of 1,000 scenarios, we can say that the probability of success is 80%.

As a benchmark, a probability of success of 75% or more is good.

Rates of return fluctuate from year to year and are based on the expected return and volatitlity (as measured by the standard deviation) of each asset class.

The chart below shows sample simulations for somone age 45, retiring at age 65 and with a life expectancy of 90. The variability of outcomes results from the degree of volatility, which increases with the proportion of assets invested in equities.

Why is it useful?

The goal is to raise the comfort level knowing the odds of achieving pragmatic financial lifestyle goals, and feel comfortable with a financial plan, even in periods of negative market performance.

If a probability of 80% percent seems too risky and you would like to increase the odds to 85% or 90% , you can adjust your annual savings, retire later, reduce your income requirements or change your portfolio allocation.

The risk of hitting a string of bad years early on can easily upset a retirement plan. For instance, if you retired during a period as bad as the stock-market returns of the mid-1970s, you would run out of money very fast.

What traditional planning ignores is the timing of the returns. A Monte Carlo simulation highlights some of the problems that might arise in a down market shortly after retirement.

If there is not a high enough probability for success in achieving retirement goals, changes such as retiring older, saving more, adjusting income expectation or a combination of these become clear.

Success in early retirement can set the tone

It is important never to lose sight that the type of investments and allocation within a portfolio have a direct impact on the amount of volatility that can occur. Small foreign companies will have large amounts of variation in returns, while high-grade short term Government will have very little.

Monte Carlo should not be viewed as a certainty test. It is a probability test. Ultimately, there will be only one outcome, but knowing not to take more risk than necessary and finding a safe spending level is invaluable information.

By: RetireWare / Equisoft On 4:10 PM

Continue Reading