In 30 Seconds ChatGPT Casts Doubt On Minnesota Pension 30 Year Performance "Miracle"
Artificial Intelligence Agrees Minnesota State Pension's Performance Claims Look Suspect
The $140 billion State of Minnesota pension fund claims to have beaten the market regularly over the past 30 years— by the exact same percentage, 0.2.
That’s remarkable, exceptional…. even unbelievable.
As governor, Tim Walz, the Democratic Vice Presidential candidate was Chairman of the Board of the pension. The pension performance claims are even more remarkable given that Walz admits he has never owned a stock or bond. Walz may have missed his calling, given that he’s outperformed virtually all Wall Street professional money managers. He should be shouting about these investment results….but, he’s not. Neither is anyone else in Minnesota.
Thanks to behind-the-scene pressures applied by state politicians, Minnesota media has also been mum about, i.e., has completely ignored, the performance claim controversy surrounding the state pension—despite the impact the enormous fund has on the lives of all Minnesota pension stakeholders, including pension participants and taxpayers.
Nevertheless, in Ohio, the Minnesota state pension performance claims recently caught the attention of the Editor of the Toledo Blade:
“A long term look at Minnesota’s pension math is just as perplexing. The teachers retirement fund purports to beat a composite index they created by 0.2 percent measured over 1, 5, 10, 20, and 30 years. The odds of that level of consistency over each measure of time are infinitesimal (emphasis added).”
The odds of that level of consistency over each measure of time are infinitesimal- Toledo Blade
So, last night we asked ChatGPT to weigh in on the issue:
What is the likelihood that an investment fund could beat the market by exactly the same number: .02% on a 1, 5, 10, 20 and 30 year basis?
30 Seconds later ChatGPT said the likelihood was extremely low and offered the following analysis:
The likelihood that an investment fund could beat the market by exactly the same margin (0.02%) over different time periods (1, 5, 10, 20, and 30 years) is extremely low (emphasis added). Here’s why:
Market Variability: Markets are inherently volatile, and the performance of an investment fund relative to the market can vary greatly over time due to changes in economic conditions, interest rates, market sentiment, and other factors. Consistently outperforming the market by the same margin over different time periods would require an extraordinary level of stability and predictability, which is rare.
Statistical Unlikelihood: The probability of achieving the exact same difference in performance over multiple time frames is very low due to the statistical nature of investment returns. Small differences in returns are often influenced by a variety of factors that can change over time, making it unlikely to see the same exact margin of outperformance over different periods.
Performance Measurement: Fund performance is typically measured with varying degrees of precision depending on the time frame. Short-term performance (like 1 year) can be heavily influenced by short-term market fluctuations, while long-term performance (like 30 years) is more reflective of the fund’s underlying investment strategy and market conditions over an extended period. The exact margin of outperformance can be influenced by many factors that may not remain constant.
Luck vs. Skill: Achieving exactly the same outperformance margin across various time periods could be indicative of either extraordinary skill or an element of luck. In practice, even skilled fund managers find it challenging to maintain consistent outperformance over different time horizons due to the dynamic nature of markets.
In summary, while it’s theoretically possible for a fund to beat the market by the same margin over different periods, the likelihood of this occurring is extremely low due to the inherent variability in market performance and the numerous factors affecting investment returns.
We then asked ChatGPT:
Are the benchmarks public pensions use to gauge their investment performance biased?
To which, ChatGPT responded:
Public pension benchmarks can sometimes exhibit biases, and understanding these biases is crucial for evaluating the performance of public pension funds accurately. Here are several ways in which benchmarks might be biased:
Selection Bias:
Index Composition: Public pension benchmarks often use indices that may not fully represent the fund's actual investment universe. For example, if a benchmark is based on a broad market index, it might not account for specific asset classes or sectors that the pension fund holds, potentially skewing performance comparisons.
Survivorship Bias: If a benchmark includes only funds or securities that have survived over a long period, it might exclude those that have underperformed or failed. This can result in an overly optimistic view of performance.
Performance Bias:
Target Setting: Benchmarks might be set based on historical performance or specific targets that don’t necessarily reflect the current investment environment or the fund’s strategic asset allocation. This can make it challenging to compare the fund's performance accurately.
Custom Benchmarks: Some pension funds use custom benchmarks designed to match their specific investment strategy or risk profile. These benchmarks can be crafted in ways that make performance look better than it might be when compared to more standardized benchmarks.
Time Period Bias:
Short-Term vs. Long-Term: Benchmarks might be evaluated over different time periods than the fund’s performance periods. Short-term benchmarks may not capture long-term performance trends, and vice versa, potentially skewing the perceived success of the fund.
Asset Allocation Bias:
Strategic Allocation: A benchmark might not reflect the actual strategic asset allocation of a pension fund. If a fund has a unique allocation strategy that deviates significantly from the benchmark, the comparison might not be fair or meaningful.
Sector and Style Bias:
Sector Weighting: Benchmarks might be heavily weighted in certain sectors or asset classes, which can create biases if the pension fund has a different sector allocation. For instance, a benchmark heavily weighted towards technology might not accurately reflect the performance of a pension fund with a more diversified or conservative allocation.
Risk Profile Bias:
Risk Tolerance: Benchmarks might not align with the risk profile of the pension fund. If a benchmark is more volatile or less volatile than the fund’s investments, it can create misleading comparisons regarding performance.
Mitigating Biases: To address these biases, it’s essential to:
Choose Appropriate Benchmarks: Use benchmarks that closely match the pension fund’s asset allocation, risk profile, and investment objectives.
Regularly Review Benchmarks: Periodically review and adjust benchmarks to ensure they remain relevant and reflective of the fund’s strategy.
Use Multiple Benchmarks: Consider comparing performance against multiple benchmarks to get a more comprehensive view of how the fund is performing relative to various market segments.
Understanding and accounting for these biases helps in making more informed evaluations of a pension fund’s performance.
End ChatGPT
In conclusion, Minnesota state pension financial statements raise immediate, obvious, serious concerns. As the Toledo Blade summarized:
“A cursory look at the Minnesota Teachers Retirement Association leads to the conclusion they’re either a world class pension or they’re cooking the books.”
It’s clear that for the past 30 years, public understanding and scrutiny of the workings of the Minnesota state pension system has been sorely lacking. Equally clear, state officials and their Wall Street allies don’t want anything to change.