A client recently came to me with a retirement portfolio allocation recommendation from their large financial institution. They charge 1% a year to manage the client's retirement portfolio. With nearly $2,000,000 in retirement assets, that's almost $20,000 a year.
But for that steep price...did they give them good advice?
Spoiler alert: No. I'll show you why below.
As part of my comprehensive wealth plan, I offer clients asset allocation advice in addition to estate, wealth, and retirement planning. If you hire me for this service, you can rest assured that my advice is exclusively for your benefit. That is because I have no targets to meet, no quotas to fulfill, no centrally dictated guidelines to enact. Rather, I work exclusively for my clients.
However, if your current advisor works for a large institution, they operate on a tight leash. Often their "custom" recommendations emanate not from them but some back office somewhere. They put you and your dreams in whatever square hole they have at their disposal - regardless of whether it is an appropriate fit.
As a Life and Legacy attorney in Austin, Tx, that is not how I operate. Keep reading, and I'll show you how I approached this retirement portfolio allocation question.
If you find this post interesting, I encourage you to check out some interesting opportunities for further enrichment on the resources section of my website.
I want to show you how I approached this problem. For a fraction of the cost the client was paying, I believe that I provided better advice. Not only will I save this client paying $20,000 a year for what I believe is bad advice, I also provide them with the opportunity to perhaps earn higher returns for less risk.
As shown below, the combination of these two enhancements could potentially add up to a considerable sum at the plan's conclusion.
| Asset Class | Status Quo Allocation | Recommended Allocation |
| Guaranteed | 0.0% | 14.9% |
| Equities | 55.59% | 56.4% |
| Large-Cap Growth Stocks | 8.73 % | 8.2% |
| Large-Cap Value Stocks | 8.30 % | 9.7% |
| Large-Cap Blend | 19.36 % | 0.0% |
| Mid-Cap Growth Stocks | 0.0% | 4.5% |
| Mid-Cap Value Stocks | 0.0% | 4.0% |
| Small-Cap Growth Stocks | 0.0% | 3.4% |
| Small-Cap Value Stocks | 0.0% | 2.6% |
| Small-Cap Blend | 7.96 % | 0% |
| International Stocks | 7.59 % | 15.9% |
| Emerging Markets | 3.62 % | 5.0% |
| Real Estate Securities (REITs) | 0.0% | 3.1% |
| Real Estate | 0.0% | 8.1% |
| Direct Real Estate | 0.0% | 8.1% |
| Fixed Income | 44.41% | 19.1% |
| High Yield Bonds | 0.0% | 3.1% |
| Inflation-Protected Bonds | 17.06 % | 3.6% |
| Bonds | 16.34 % | 2.7% |
| Long Term Bonds | 3.97 % | 3.4% |
| Short Term Bonds | 3.53 % | 3.5% |
| Foreign Bonds | 0.0% | 1.1% |
| Long Term Muni Bonds | 0.0% | 0.6% |
| Short Term Muni Bonds | 3.51 % | 1.1% |
| Cash/Money Market | 2.2% | 1.5% |
The Status Quo allocation has an estimated rate of return of 5.95% and a standard deviation of 6.31%. The $20k allocation has an estimated rate of return of 4.78% and a standard deviation of 9.26%.
So, the recommended allocation has a lower rate of return and a higher standard deviation: the opportunity to make less money for more risk.
But, has the return per unit of risk decreased? Let's compare the Sharpe ratios:
The Sharpe ratio is merely the expected portfolio return minus the risk-free rate over the expected standard deviation of the portfolio. The risk-free rate today is about 1.52%
So, the Sharpe ratio for the Status Quo retirement portfolio allocation is 0.70, for the $20k allocation recommendation is 0.35. Both are terrible. Below 1 is unacceptable. Above 1 is good; above 2 and 3 are excellent. But, the $20k recommendation has a Sharpe ratio of half the current allocation.
Given the poor Sharpe ratio of the Recommended allocation, I will dismiss it outright (I will only return to it if I cannot beat the return and standard deviation through our optimization models below).
But, given that this is a retirement portfolio allocation problem, there may be limits to how high we can get the Sharpe ratio. That is because there is an understandable aversion to risk exposure for a retiree. The preferred allocation for this retiree is between 50-60% exposure to equities.
However, perhaps we can do better through portfolio optimization nonetheless.
First, we will find the historical returns of the investment funds. Second, we will calculate the covariances of the funds to see how they move relative to each other. Third, we will optimize the portfolio to maximize return for a limited variance and a minimum variance for a specified minimum return. In both cases, we will look to maximize the Sharpe ratio as well.
First, let's see if we can improve the Status Quo retirement portfolio allocation's return and standard variation through better allocation. We are trying to see if we have the opportunity to make more money with less risk.
The first optimization has the objective of minimizing the variance. A requirement also constrains the problem that exposure to equities does not exceed 60%. Further, the total portfolio allocation must equal 100% with no shorting (i.e., no negative allocations). Finally, we set the desired return to greater than or equal to 9%.
Our model returns an optimized retirement portfolio allocation with an expected return of 9.17% and an expected standard deviation of 6.84%. Furthermore, the Sharpe ratio climbs to 1.119, so in the good range.
Moreover, this optimized portfolio also benefits from simplification over either the current or recommended portfolios above. Specifically, the optimization model recommends about 1/4 of the assets allocated to both an S&P 500 index fund and a bond fund. The balance is allocated about 2/3rds to a large-cap growth stock fund and the final 1/3rd to a TIPs fund.
How does the optimized retirement portfolio allocation perform relative to the Status Quo allocation? We test both portfolios over the prior 10 years according to their respective allocations.
Starting with $2,000,000 at the beginning of 2012, how much would you have at the beginning of July 2021?
| Year | Status Quo Portfolio Returns | Status Quo Portfolio Balance | Optimized Portfolio Returns (MinVar) | Optimized Portfolio Balance (MinVar) | Difference in Balance |
| 2012 | 7.78% | $2,034,059.65 | 7.94% | $2,158,806.55 | $124,746.90 |
| 2013 | 7.02% | $2,176,860.01 | 10.98% | $2,395,883.32 | $219,023.31 |
| 2014 | 4.56% | $2,276,212.65 | 7.87% | $2,584,406.90 | $308,194.25 |
| 2015 | -6.70% | $2,123,699.77 | -3.08% | $2,504,830.78 | $381,131.00 |
| 2016 | 10.79% | $2,352,918.97 | 9.99% | $2,755,123.18 | $402,204.20 |
| 2017 | 12.11% | $2,637,899.84 | 17.05% | $3,224,937.23 | $587,037.40 |
| 2018 | -5.23% | $2,500,062.82 | -2.89% | $3,131,797.61 | $631,734.79 |
| 2019 | 9.76% | $2,744,095.72 | 15.84% | $3,627,848.95 | $883,753.23 |
| 2020 | 10.87% | $3,042,434.80 | 16.42% | $4,223,431.19 | $1,180,996.39 |
| 2021 (YTD) | 8.50% | $3,301,071.73 | 11.59% | $4,712,773.24 | $1,411,701.51 |
So, if you had invested the $2,000,000 portfolio according to the optimized allocation, after almost 10 years, you could have had an additional $1.4m in your portfolio!
In every year, except 2016, the optimized portfolio either performed better or less worse than the current portfolio.
That is, more return for less risk!
The second optimization model seeks to maximize return for a variance of less than 1%. The model returns an expected return of 10.01% and a standard deviation of 8.32%. The Sharpe ratio is 1.02. The optimized allocation, in this case, is split into just two funds: about 3/5ths into a large-cap growth fund and 2/5ths into a bond income fund.
Given that allocation, the backtest from 2012 to YTD 2021 with an initial portfolio of $2m results in $5,039,555.12.
| Year | Status Quo Portfolio Returns | Status Quo Portfolio Balance | Optimized Portfolio Returns (MaxRet) | Optimized Portfolio Balance (MaxRet) | Difference in Balance |
| 2012 | 7.78% | $2,034,059.65 | 9.46% | $2,189,160.05 | $155,100.40 |
| 2013 | 7.02% | $2,176,860.01 | 12.06% | $2,453,216.73 | $276,356.71 |
| 2014 | 4.56% | $2,276,212.65 | 7.15% | $2,628,527.76 | $352,360.11 |
| 2015 | -6.70% | $2,123,699.77 | -5.71% | $2,478,481.34 | $354,781.57 |
| 2016 | 10.79% | $2,352,918.97 | 11.99% | $2,775,651.35 | $422,732.38 |
| 2017 | 12.11% | $2,637,899.84 | 18.96% | $3,301,970.62 | $664,070.78 |
| 2018 | -5.23% | $2,500,062.82 | -3.41% | $3,189,241.32 | $689,178.51 |
| 2019 | 9.76% | $2,744,095.72 | 18.06% | $3,765,218.44 | $1,021,122.72 |
| 2020 | 10.87% | $3,042,434.80 | 20.06% | $4,520,672.05 | $1,478,237.25 |
| 2021 (YTD) | 8.50% | $3,301,071.73 | 11.48% | $5,039,555.12 | $1,738,483.39 |
The difference between the status quo portfolio and the maximized return portfolio in the backtest from 2012 to 2021 (YTD) is $1,738,483.39!
Sure, going from backwards to the present is a good first sanity test. But, how can we test how the portfolios might perform going forward?
Using the standard deviation and expected returns, we can randomly draw returns for each projected year. In addition, we can do that many times and then look at different statistics, such as the mean and minimum. Finally, we can also look at the percentage chance the final portfolio will fall below a certain threshold (i.e., the chance the final portfolio is less than $100,000?). This is a monte carlo simulation.
To keep this post from going too much longer, I will only compare the Status Quo portfolio with the Minimum Variance optimized portfolio.
Starting with the Status Quo portfolio, we run our monte carlo simulation 10,000 times.
After 24 years of retirement, taking annual required minimum distributions, the Status Quo portfolio has an average final value of almost $850k with a standard deviation of $250k. So, the likely final value of the portfolio is between $600k and $1.2m.
A 76% chance exists in this model that the final value will fall below the retiree's wish to leave a $1m bequest.
Next, using the optimized Minimum Variance retirement portfolio allocation, we run our monte carlo simulation another 10,000 times. This is like our retiree having 10,000 tries at retirement, each year of returns over 24 years assigned randomly within a normal distribution of an expected return of 9.17% and a standard deviation of 6.84%.
Therefore, starting with a $2m portfolio, our retiree allocates according to the model's optimized minimum variance recommendation. They also take required minimum distributions (RMDs) out annually. Hence, at the end of 24 years, they have a mean of $1,7m left in the account, with a standard deviation of about $550,000. So, they average between $2.2m and $1.2m to leave as a bequest. There is only a 4% chance that they have less than $1,000,000.
Finally, we ran the above monte carlo simulation with the assumption that the returns are normally distributed. That isn't the case, as the graph below shows. Instead, the returns are distributed according to a minimum extreme distribution.
Next, rerunning the monte carlo simulation with the fitted distribution, we actually get rather similar results. The mean potential bequest is just over $1.7m, with a standard deviation of about $575k. And, there is a 7% chance of there being less than $1m after 24 years of retirement (and taking the annual RMDs). So, not much difference.
Nonetheless, I find that it is always useful to do a sanity check when performing such calculations. Although much of financial modeling depends on normal distributions, actual finance does not always perform "normally."
So, our retiree was spending $20,000 a year for bad advice. Over 20 years, that's nearly $500,000. In addition, we saw that a portfolio with an allocation optimized to either minimize variance or maximize return had better Sharpe ratios. And they returned a potential mean expected return of between $1.4m and $1.7m more than the Status Quo portfolio, respectively.
Altogether, by investing about 1/5th of the fees she was paying for poor advice in my services, this retiree is on track to gain perhaps $1.9m to $2.2m relative to the Status Quo or Recommended portfolio allocation.
I, of course, cannot guarantee you the same results with your retirement portfolio allocation. But, I can guarantee you that I will carefully and thoroughly analyze your portfolio allocation and optimize it according to your individual needs if engaged in doing so.
In conclusion, if you will spend $20,000 on advice with a big financial institution, at least make sure it isn't going to waste.
(By Appointment Only)
14425 Falcon Head Blvd
Bldg E-100
Austin, TX 78738
