The Risk Parity Saga Continues

November 10, 2026 – Risk Parity has been one of the most requested topics for years. I always pushed back, finding the approach silly, inefficient, and thus useless for retirement portfolios. But finally, in July of 2026, I published two blog posts: one on the safe withdrawal rate impact of Risk Parity (SWR Part 64) and one more general post about the lies and misunderstandings in Risk Parity, Part 4 in my “How to Lie with Personal Finance” (tongue-in-cheek!) Series. I thought that was it, and I would leave it at that. But I certainly stepped on somebody’s toes: my old financial influencer buddy Frank Vasquez wasn’t happy. He issued his reply recently. I thought long and hard about whether I should even reply, but in the end I wrote some notes that filled an entire blog post. So, let’s take a look and put this crazy Risk Parity idea to bed, once and for all…

Caution: this post contains sarcasm, hyperbole, and figures of speech. It also contains numbers, calculations, and cold hard facts. This may not be suitable for the faint of heart!

Frank Vasquez has replied in Episode 548 of his podcast. You may wonder what I think about his innuendo, anecdotes, red herrings, and pettiness. I must be living rent-free in that guy’s head. I don’t dignify that with a reply. Instead, let me address the replicable, quantitative, and indisputable issues. Let me go through ten responses to Frank’s recent work:

1: The terrible live performance of Risk Parity post-2020

Frank never addressed my main concern in his reply: he can generate all the nice-looking simulations he wants (which I will chop down below as well, by the way, stay tuned), but he can’t defend the terrible recent performance of his Risk Parity sample portfolios. So let me rub in the inferior performance of his recommended Risk Parity portfolios once more, this time in more detail.

First, let me take a step back and state the obvious. It’s one thing to point out how well your portfolio did in the past. I can play the same game and easily beat Frank by adding a large enough allocation of QQQ to my simulations. Or NVDA, for that matter. And then claim I’ve solved retirement math because I found a 10% safe withdrawal rate. Nobody would take that seriously because of hindsight bias. 

It’s much harder to keep your credibility if you put your portfolio to the test with real money and in a real retirement. I have consistently endorsed a 75/25 portfolio as my baseline for both my personal early retirement and in my research. This portfolio survived all historical worst cases reasonably well. It was never the best, but it was also never the worst. 75/25 is simple and avoids overfitting, and every investor understands it! It has also done reasonably well since my retirement in June 2018, and after the 2020 and 2022 market peaks, all with live performance.

Not so much Frank’s Risk Parity portfolios; please find the return data and more details here, updated to 9/30/2026. He started most of his sample portfolios in late July 2020, and the performance has been atrocious so far. The 75/25 portfolio would have mopped the floor with every one of his Risk Parity portfolios; see the table below. And yet, he’s touring the podcast scene and telling people that Risk Parity all but guarantees better retirement performance, with 5% safe withdrawal rates. That’s false advertising:

• Better portfolio? Totally false. The 75/25 portfolio vastly outperformed all the Risk Parity portfolios. It outperformed the best Risk Parity portfolio by almost 4 percentage points. Some of the other portfolios are actually so bad, they didn’t even get a positive nominal return (Aggressive 50-50), and two more failed to have positive real returns (All Seasons, Accelerated PP), so even before withdrawals, you’d be underwater with these trash portfolios, despite one of the best bull markets in history.
• Less Risk? Partially true: some Risk Parity portfolios had lower risk, but at the cost of far lower compounded returns. The 75/25 portfolio had by far the best risk-adjusted returns (Sharpe Ratio) at 0.69, compared to the Risk Parity portfolios (between a pathetic -0.15 and a so-so +0.49). Which is funny, because Frank always claims he’s working with the tools of Harry Markowitz, who’s rolling in his grave at how uneducated podcasters are bastardizing his classic portfolio optimization framework today.
• Shallower and shorter drawdowns? As he claimed here (~43:27 timestamp), here (~22:28 timestamp), and here (~44:32 timestamp). Not true. Quite the opposite: 75/25 had a shallower drawdown (even before withdrawals) than all Risk Parity portfolios except the Golden Butterfly (less than a 2-point difference, though), and it had the shortest drawdown of all in the 2022 bear market. Worse, three of the portfolios (All Seasons, Accelerated PP, and Aggressive 50-50) still had ongoing drawdowns in real-CPI-adjusted terms, even without withdrawals! And everything looks far worse with 5% annualized withdrawals, which leads me to the next point…
• A sustainable 5% withdrawal rate in retirement? Not true. If you started in July 2020 and withdrew an annualized 5% from your initial portfolio (then CPI-adjusted), only the 75/25 portfolio performed well, reaching a new all-time high in September 2026 (+11% CPI-adjusted after withdrawals). The Risk Parity portfolios are an utter clown show, with many well below their starting values. Even the better ones are 14% and 17% below their starting value in 2020, which means that if you insist on withdrawing 5% of the initial portfolio value, you now withdraw 5.83% (GB) and 6.05% (GR). The bad ones are significantly compromised. If you continued withdrawing 5%, you would run out quickly, especially in the remarkably leveraged portfolios with a 20%+ realized risk. (Note: I assume a WR of 5% for all portfolios, except the All-Seasons portfolio, which assumes 4%.)

2: Real vs. Nominal Risk Parity Shenanigans

Wait a second, how is it possible that Frank is always touting that his Risk Parity portfolios are up, when my table above shows they are all down, and most of them even dangerously depleted and a few years away from bankruptcy? Well, Frank muddies the water by presenting nominal figures, not real figures. So technically he’s correct in some cases, but in real, CPI-adjusted terms his portfolios look disappointing.

Side note 1: If you wanted to gather the information I provided in the table above on Frank’s website, you’d be hard-pressed. Frank doesn’t provide it. Frank used to post tables with monthly returns. I captured a few screenshots back in the day, but Frank was wise enough to stop posting such detailed data because the returns were so atrocious. How sneaky! Again, if you want to see the returns, return stats, and screenshots, see here.

Side note 2: Some of the portfolios are down even in nominal terms; see the numbers as of October 30 below:

In any case, yes, the Golden Ratio portfolio balance may look great, at $11,783 as of October 30, 2026, compared to $10,000 in July 2020. However, it’s actually down if you adjust for inflation. I don’t have the October CPI number yet, but if I even take the September 2026 CPI index (336.298) and compare it to the July 2020 figure (258.543), then the portfolio value in July 2020 is $9,059, or almost 10% below its starting value. All the other portfolios also perform poorly by this measure. For example, the Aggressive 50-50 portfolio is down to $6,696.43 in nominal dollars on October 30, 2026. $5,148 in real dollars. Ouch! Almost 50% depleted after six years. Does Frank believe this portfolio will make it through the next decade?

The same shenanigans also apply to Frank’s withdrawals. He must have withdrawn significantly less than what I assumed in my simulations, i.e., 5% of the initial portfolio, CPI-adjusted! Let’s dig into the details. Frank claims he pulled various initial withdrawal rates, ranging from 5% to 8%, from his portfolios starting in July 2020 (one exception: the Levered GR portfolio started in July 2021), adjusted for inflation. But the total withdrawals he reports for CY 2025 seem awfully small by comparison. In the table below, I compare his strategies, compute the initial withdrawal, and compare the target CY 2025 withdrawal under a CPI-adjusted profile with his actual CY 2025 withdrawals.

Example: for the All Seasons Portfolio, he withdrew $10,200 x 0.04 / 12 = $34.00 in the first month (July 2020). If we adjust this amount for CPI every month, we should expect $483.45 in 2025. The actual withdrawals were only $384 (I took a screenshot on his site in early November), which is much lower (-20.6%). In fact, his 2025 withdrawals were only 3.18% of the initial portfolio, adjusted for CPI. Likewise, every strategy’s 2025 withdrawal amount was much lower than the initial amount, adjusted for inflation. In hindsight, Frank would have been far better off with the 75/25 portfolio, which would have fared all right with a 5% initial withdrawal rate. Though, without the benefit of perfect foresight, I would not have endorsed that as particularly safe back in 2020.

How is it possible that he tightened his belt by between 5% and 58% since he started his withdrawals? Very simple: you must read the fine print. On his blog, all his portfolios have a qualifier. For example, all portfolios with a 5% safe withdrawal rate have this line:

“If the portfolio falls below 80% of its starting value, we will reduce the withdrawal rate to 4% until it returns to its starting value.” Source: https://www.riskparityradio.com/portfolios

Specifically, every portfolio has this lower (flexibility-adjusted) withdrawal rate between one and two percentage points below the initial rate. So much for the safe withdrawal rate of 5%; it can suddenly turn into 4%. I don’t remember him ever mentioning this on the podcasts where he touts his 5% safe withdrawal rate and superior retirement performance relative to the fixed withdrawal of other portfolios. Gee, I wonder why?!

So, to sum up, all of the Risk Parity portfolios also have sizable declines in spending power. The Golden Butterfly (-5%) and Golden Ratio (-9%) are the least impacted. The leveraged portfolios are a complete disaster (-44.6% to -58.3%), and the two other portfolios are only moderately awful (-20.6% and -34.7%).

3: Fixed vs. Variable Rate Shenanigans

Looking at the table again, notice that the actual withdrawals as a percentage of the initial portfolio dropped well below not just the initial percentage withdrawal rate, but even the lower/flexible percentage rate. The four portfolios, RP Ultimate, Accelerated PP, Aggressive 50-50, and Levered GR, all have initial rates between 6% and 8% and flexible rates between 5% and 6%. But all are now withdrawing between 3.33% and 4.03% p.a. of the initial portfolio. Why are these figures lower than 5% and 6%, respectively? Very simple: there is more fine print! For example:

“We will be distributing 8% annualized from this portfolio each month, which is the mathematical equivalent of dividing the ending monthly balance by 150 and distributing that amount on the first trading day of the next month. If the portfolio falls below 80% of its starting value, we will reduce the withdrawal rate to 6% until it returns to its starting value.” Source: https://www.riskparityradio.com/portfolios

Well, isn’t that cute? The 8% doesn’t refer to the usual fixed-withdrawal-rate arithmetic that Bengen, Trinity, and yours truly use with my SWR toolbox. The 8% is simply a variable percentage withdrawal rate, essentially a dumbed-down version of the Bogleheads VPW (without the amortization feature, but with a much higher % rate). Well, that’s very deceptive! For example, you could have withdrawn about 7% annually from your 100% equity portfolio and maintained your real equity balance over the last 150+ years, because 7% was the average, CPI-adjusted compound return for the S&P 500. But do you see the drawback, though? This is not a flat spending path. Your spending is just as volatile as your portfolio. You should never compare this highly volatile VPW spending rate to a Bengen/Trinity/ERN-style fixed withdrawal rate. It’s apples and oranges – False Advertising!

Now, some readers may ask, what’s the problem? Frank’s favorite Golden Ratio portfolio now withdraws only 9% less than initially. Only? Ask the average American, maybe someone on Social Security or on a pension only, how things are going. They will tell you that they are getting crushed in this decade. And they are getting an annual COLA in their Social Security. Now try the same retiree with COLA minus 1.8% annualized (1.8% = 9% spread over 5 years), and you’re welcome to Frank’s “tighten your belt” retirement club! That’s a disaster in retirement. And let’s not even talk about the poor folks who might have followed Frank’s insane alternative portfolios, where your spending is down between 44% and 58%.

Talking about spending. This is an outrageous quote from his Episode 548:

“The last time I heard him talk about his personal situation, which is this summer, he said they haven’t been spending their money. They spend about 2% out of the portfolio. He doesn’t sell anything. Number go up is a fundamental part of his identity. That is how he values himself. Now, has he increased his spending? Well, yes, but guess what? It didn’t have anything to do with inflation. It was just a choice. He’s basically created a variable spending plan for himself. If his numbers keep going up enough, he’ll spend a little more money. So while his portfolio has gone up two and a half times or something like that, his spending has gone up 75%. In my mind, he should have been spending 75% more at the beginning.” Source: Episode 548, RiskParityRadio.com

He has no clue about my spending patterns. Creepy Uncle Frank must be hallucinating! He’s also accusing me of running a variable spending plan, when that’s exactly what he is doing, documented on his own page. Projection! He probably had to cut his own spending because the portfolio was down. Maybe this is why Frank is so cranky all the time. I’m much better off; since retiring, I’ve raised our budget by about 2.7x nominally and about 2x in real, CPI-adjusted terms (CY 2025 vs. the first year). We splurge and spend lavishly; earlier this year, my wife and I spent the equivalent of our entire initial annual retirement budget in one single day. Frank is probably jealous.

If you thought Frank’s July 2020 retirement cohort was bad, wait until you see the January 2022 cohort! This brings me to the next point…

4: Not calculating a true safe withdrawal rate for Risk Parity

My 75/25 portfolio would have supported a 7.37% withdrawal rate since my retirement in June 2018, meaning it would now be back to its CPI-adjusted initial value (not even counting my deferred bonuses, options trading, blog income, etc.). But I’m not touting my strategy as a safe 7% withdrawal rate. As a serious retirement safety researcher, I gauge a safe withdrawal rate by historical worst-case scenarios.

Frank, in contrast, knowingly and intentionally ignores the definition of (fail-)safe withdrawal rates. Obviously, he can and should study how his portfolio would have performed since July 2020, or from earlier start dates, limited by the return availability on testfol.io. But the real test for his portfolios comes from the historical worst-case scenario. That was not the month Frank started his simulations (about July 2020) or the month return data for the Golden Ratio portfolio became available (January 2000), but likely the market peak on 12/31/2021. 

So, let’s assume one of Frank’s listeners was so gullible and unfortunate as to start retirement on 1/1/2022 using a 5% withdrawal rate and one of his crazy portfolios. Let’s check the table below. If you had used even the better-looking portfolios (Golden Butterfly and Golden Ratio), your portfolio value is now depleted to only 83% and 78%, respectively, of the initial. If you keep withdrawing 5% of the initial portfolio value adjusted for inflation, you’d now withdraw 5.83% and 6.05% in the GB and GR portfolios, respectively. RP Ultimate and Levered Golden Ratio are each down 30% and 32%, respectively. The other portfolios are down 50% (Accelerated PP), and even 67% (Aggressive 50-50) if you had used a 5% initial fixed withdrawal rate.

Maybe if you had started with a 30-year horizon and you’re now down to about 25 years of life expectancy, you might still make it with the Golden Butterfly and Golden Ratio. But for early retirees, say with a 50-year initial horizon, a withdrawal that high coming out of low-average-return portfolios will almost certainly fail in the remaining 45 years. Some of the other portfolios (Accelerated PP and Aggressive 50-50) are so devastatingly depleted that they will not even make it another 10-15 years with 5% withdrawals.

By the way, the 75/25 portfolio, while again performing the best among all, is also down to 93% of its initial value. I would not keep withdrawing 5% from that portfolio in today’s world of high earnings multiples. But again, I’m not a ditzy podcaster, so I would have obviously never recommended a 5% SWR in the first place. However, at a 4% withdrawal rate, the 75/25 is essentially back to its initial value (99% of the initial value, adjusted for CPI inflation) and will likely do reasonably well going forward. Maybe take withdrawals down to 3.75% given the high valuations and the political and financial uncertainty ahead, though. It seems none of Frank’s portfolios are safe for early retirement, even with a 4% initial withdrawal rate, for this cohort.

In light of this complete clown show of Risk Parity portfolios and their CPI-adjusted withdrawals getting ravaged, I need to ask myself: how come the Risk Parity portfolios were safe in some of the historical worst-case scenario simulations, 1929, 1937, 1960s, and 1970s, even at a 5% fixed rate without ever having to reduce your spending? But a garden-variety market blip like 2022 sent his lineup of portfolios into a tailspin, where just 5-6 years into retirement you needed to cut spending by 5% to 58% in 2025? Most people don’t even remember that bear market anymore, because it was one of the mildest on record, and it wasn’t even accompanied by a recession. What explains the poor performance despite the incessant claims that Risk Parity should prepare your portfolio for all possible economic regimes? Very simple. Let me summarize it again:

Risk Parity portfolios were heavily overfitted to look good in historical episodes. But when running the strategy in real time, it fizzled. Some of his initial safe withdrawal rates were also too aggressive.

Why does this matter? Frank’s portfolios didn’t merely slightly underperform the 75/25 portfolio during a record bull market. It’s not like 75/25 is up by 400% and Risk Parity is up by “only” 399%; I wouldn’t complain in that case. The reality for Frank’s Risk Parity portfolios is very dire: the current portfolio values are potentially fatally underwater for the worst-case-scenario cohort as of 1/1/2022. Even the slightly better cohort from July 2020, when Frank started this project, is taking on water, and this required painful cuts to the retirement budget that will likely not improve anytime soon. A Sequence Risk nightmare!

5: Backfilling past return data

The Golden Ratio portfolio started with a 10% allocation to Global REITs (ETF ticker REET). In December 2024, Frank swapped that fund with a “Managed Futures” ETF (ticker DBMF). There is nothing wrong with that; sometimes you change/replace asset classes. But use caution when trusting his simulated performance. That’s because on his site he’s now linking a simulation that uses DBMF for the entire simulation horizon, not just post-2024. But if you had retired at the market peak in December 2021, you would have still followed Frank’s advice back then to go into REET, not DBMF. And until the switch in December 2024, DBMF would have massively outperformed, by a whopping 38% (which is why Frank swapped the ETFs, duh!). See the two portfolios side by side here: REET ended down at $8,598.35, while DBMF ended up at $11,869.21. If you plug the two different funds into your Golden Ratio retirement portfolio, starting on 12/31/2021, with 5% annualized withdrawals, then the DBMF-based portfolio would have outperformed the REET-based portfolio by a full 5.26% by 12/31/2024. So unsuspecting listeners who look up the historical simulations may not realize the simulated series on testfol.io looks about 5% better than what Frank actually achieved over those three years.

Even ignoring the ETF swaps, there is more manipulation. Here’s a screenshot from his site explaining the Golden Ratio portfolio:

When you click the testfol.io link on his page, you will find this parameterization. Do you notice something? The equity funds are different. So the simulations don’t match Frank’s actual portfolio.

Gee, I wonder if Frank might have screened for better-performing ETFs to (retroactively!) replace his underperforming ETFs? So let’s check how Frank’s current portfolio would have performed side-by-side with the one he’s using in the simulations:

And below is the performance of the two portfolios on testfol.io. As I expected, the simulated portfolio with VIGRX and DFSVX outpaced Frank’s actual portfolio with VUG and VIOV. By about 7.5%, i.e., $9,427.18 vs. $8,798.60 final CPI-adjusted value.

So, be cautious with Frank’s simulations. He has manipulated the simulations in at least two ways: 1) He replaced ETFs midway in his actual portfolio (removed REET, added DBMF) but then displays the simulations as if he’s held the better-performing DBMF throughout. 2) He uses better-performing equity ETFs for small-cap value and large-cap growth to make the simulations look better than the actual picks he introduced in 2020 and still holds today. Very deceptive! Obviously, ditzy internet influencers can do that – not much reputation loss or legal liability there. However, if you worked in any licensed and regulated function in the finance industry, you’d get a call from the regulator for that kind of simulation fraud!

6: Risk Parity ≠ Small-Cap Value

What Frank (intentionally?) misunderstands is that the Small-Cap Value discussion in the context of Risk Parity is a complete red herring. I’ve said this before, so let me repeat it here in boldface letters:

Risk Parity doesn’t have a monopoly on Small-Cap Value!

Just ask Paul Merriman, Larry Swedroe, Ben Felix, and scores of other financial influencers. Lots of people do SCV who have never heard of Risk Parity. Likewise, Risk Parity doesn’t require Small-Cap Value. Quite the opposite: the oldest Risk Parity models, Permanent Portfolio (first introduced in 1981, not true risk parity but certainly designed with the same idea) and All-Weather/All-Seasons (popularized in 1996 in the hedge fund world, 2014 in the retail world), have no Small-Cap Value allocation. The way the “godfather” of Risk Parity, Ray Dalio, sees it, the method has always been about macro betas, not stock style factors.

Moreover, if you believe that Small-Cap Value will stage a great rebound in the future – I seriously doubt it, but bear with me for now – then it’s still not a great selling point for Risk Parity. I showed in my Part 64 that you’d still use a 75/25, just with a lot of SCV in the equity portion. You wouldn’t touch Risk Parity with a six-foot pole even in that case! You most definitely wouldn’t use Frank’s Golden Ratio portfolio with only a net 4.5% HML exposure if you have true Value conviction.

I understand that Frank wants to push SCV, because attacking my stance on SCV is really his only talking point. I don’t deny that there are pros and cons of including Value and Size factors. He might even gather some support from the SCV corner of the influencer world. But that’s all a distraction. To explain this, let me use the following analogy:

Imagine you take a plate of two-day-old soggy fries you found in the trash dumpster behind a McDonald’s. Now you buy some fancy mayonnaise and some super-expensive white truffle. Not the cheap fake stuff, but the expensive ones from Piedmont, Italy. You put the Mayo and the truffle on the old fries, and they taste good again. They don’t even taste like trash anymore. Some people would call that a success. I won’t, because I’d rather put the Mayo and truffles on some nice Belgian fries, fresh from my favorite “Frites” restaurant in Ghent. Why waste good Mayo and truffles on bad fries?

Can you tell how this relates to the Risk Parity and SCV discussion? The soggy old fries from the dumpster are a Risk Parity portfolio without Small-Cap Value. Like the dumpster fries, Risk Parity has been a dumpster fire (haha, see what I did there?!) in historical simulations when excluding SCV. Examples include the Permanent Portfolio and All Weather portfolios in the 1930s and 1940s, and the Golden Butterfly without SMB and HML exposure; see my simulations in Part 34. Adding SCV, like in the Golden Butterfly and Golden Ratio portfolios, improves performance, but it’s still underwhelming. So, if you believe that SCV will rally again, you should add it to the portfolio that’s been vastly preferable, i.e., a 75/25 portfolio (really anything between 60/40 and 80/20) just like I don’t want to waste expensive ingredients on bad fries from the trash can. Shoutout to my buddy Paul Merriman, who will be happy that I likened small-cap value to Piedmont truffles. But note that this was a theoretical, academic, and hypothetical exercise, not my professional opinion or forecast.

Because some folks prefer visual information, I created the graphic below in the SWR Part 64 post to explain the mechanics: If you want to extrapolate the historical SCV alpha forward, you can increase your SWR by moving a 75/25 portfolio to 75/25 with SCV. Going from that portfolio to Risk Parity with SCV reduces your SWR again. Likewise, going from 75/25 to Risk Parity only (without SCV) makes the simulations worse. Adding SCV makes them look better again. At every step, SCV improves the outcome; Risk Parity worsens it. Friends won’t let friends invest in Risk Parity!

All of this is apparently beyond Frank’s intellectual and reading comprehension; otherwise, he wouldn’t have published that rambling podcast episode. Which is kind of intriguing, because while I can understand that he’s not very versed in mathematics and statistics, as a lawyer, albeit retired, he should at least be good at reading comprehension and logic. But maybe he wants to deflect and lead this discussion to Small-Cap Value, as a last resort because he can’t defend the indefensible trash-dumpster-fire/fries that is Risk Parity.

7: Exaggerated Fama-French weights in his Risk Parity simulations

“Most people, in fact, will not take the trouble to find out the truth, but are much more inclined to accept the first story they hear.” Thucydides. Greek historian and General.

Despite the irrelevance of Small-Cap Value I outlined above, I still like to address one important lie Frank has been spreading; the 30% SMB and 27.5% HML exposures he’s been using in his simulations (for years!) are completely wrong. They are based on an email exchange on 10/10/2022 where he had asked me for advice on how to model a portfolio significantly different(!) from his current Golden Ratio portfolio, namely 27.5% Small-Cap Value plus 27.5% large-cap blend. He didn’t mention any specific ETFs he was considering.

Being a nice guy, I wanted to give this portfolio the best possible chance in historical simulations, so I gave it the highest imaginable SMB and HML factor loadings, namely 30% SMB and 27.5% HML. That’s quite aggressive because I’m not aware of any value ETF with a 1.0 beta to the HML factor. Most value funds are around 0.5 to 0.7. But one exotic ETF (ticker RZV) has a 0.91 beta, and I graciously rounded that number up to 1.0. I treated the 30%/27.5% as neither scripture nor written in stone. It was merely an academic exercise to gauge the best-case scenario, likely not something that the typical investor would achieve. For your reference, here’s the link to the Google Sheet I shared with Frank.

In any case, do you notice something? On his website, he uses a very different allocation for his Golden Ratio portfolio: only 21% small-cap value (VIOV) instead of 27.5%. Instead of a 27.5% large-cap blend, he uses 21% large-cap growth (VUG).

Let’s look up the Fama-French factor exposures of the two funds Frank uses. You will notice that VIOV has roughly 1x SMB exposure, but its value exposure is far less than 100%, and VUG will undo almost all of it with a negative(!) HML exposure of -36%:

In fact, if I multiply the Fama-French weights by Frank’s 21% weights for each ETF, you get a net exposure of only 22.5% SMB and 4.5% HML, a far cry from the 30%/27.5% Frank is using in his simulations. 

In my SWR Series, Part 64, I used those lower SMB and HML weights, and the results are nowhere near as impressive as with the 30% and 27.5% Frank is still using incessantly and deceptively. Let’s repeat, everyone: Frank is using far more Fama-French exposure than his actual portfolio to boost the simulation results artificially. In his simulations, he’s using 1.3x the SMB and 6.1x(!!!) the HML exposure of his actual Golden Ratio portfolio.

Also note that Frank always operates on a 30-year horizon and asset-depletion framework, which is completely unacceptable to early retirees.

Side note: We can also expose another of Frank’s lies he’s been spreading on various podcasts, namely that Risk Parity has only very short-duration drawdowns. Specifically, he incessantly claims that Risk Parity has a maximum(!) of 3-4 years of drawdowns:

“The kinds of portfolios that we work with have drawdowns of between three to four years, max, not average. Max.” Frank V. on this podcast at around the 44:10 timestamp

That is a brazen lie. I’m not even referring to the live return Risk Parity portfolios, which have obviously been down much longer than four years since the 2021 peak. Rather, the historical simulations had longer drawdowns. Looking at the notorious Google Sheet I shared with him, even the overly optimistic portfolio with the exaggerated SMB and HML exposures would have had extended drawdowns (even when considering only the portfolio returns, no withdrawals yet):

• Exhibit A: 74 months between 1937 and 1943.
• Exhibit B: 72 months between 1946 and 1952.
• Exhibit C: an extended drawdown of 166 months pre-1926, i.e., between 1911 and 1925. (almost 14 years)

These withdrawals would have been even worse net of withdrawals, of course. So the 4-year maximum drawdown is an irrefutable lie.

8: Fama-French alpha is likely arbitraged away now

Despite my urgent request in my 2022 email to take the results with a grain of salt, he’s still running these raw-data simulations. He should not. My email stressed that this was a pure FIY simulation result that may look impressive for this SCV flavor in past, backward-looking historical returns. Still, it has limited relevance for investors today who are looking for forward guidance. I’ve also written about the SCV fallacy in my 2019 post mentioned in the email. In a 2024 post, I provided detailed statistical analysis with test-stats showing that over the last 40+ years, there has been no statistically significant SCV outperformance in monthly returns.

Why does alpha disappear over time? Simple: once investors find an alpha source, they arbitrage it away. It may work for a while for an elite set of investors who have the knowledge, but once a style becomes popular, the alpha fizzles. A prime example is the Black-Scholes options pricing formula. Since the late 1960s, famous mathematician and hedge fund tycoon Ed Thorp had already utilized a highly profitable variation of the formula. After the publication of the Black-Scholes and Merton papers, Thorp was able to milk it a bit longer, but the profit opportunity eventually disappeared, apparently after a decade.

I’ve made my point again in Part 62: do not extrapolate the raw Fama-French results into the future. Still, Frank uses those historical alpha returns, which are highly unlikely to repeat. And he now falsely claims that back in 2022, I supported his portfolio. I did not. Not even then (when I assumed 30%/27.5% SMB/HML), and much less now that I know what’s actually in his Golden Ratio portfolio (22.5%/4.5% SMB/HML)!

Think about it: if I had ever considered this portfolio a viable solution to increase your safe withdrawal rate, don’t you think I would have proudly announced this on my blog the next day? Don’t you think I would be aggressively marketing it in my financial advisory business now? Don’t you think I would put my own little twist on it, make it sufficiently different from Frank’s version, market it as the ultimate solution for retirement safety? Make a lot of money from it? I didn’t use it because I think pure Risk Parity is junk to begin with, and the small-cap value premium is too volatile and too sporadic to use reliably in both a 75/25 portfolio and the dumpster-fire/fries Risk Parity portfolio. Risk Parity plus Small-Cap Value is “amateur hour meets the Dunning-Kruger effect on steroids.”

9: Ignoring the pre-1929 safe withdrawal rates

Frank conveniently ignores the poor performance of his Risk Parity portfolio pre-1929. This matters regardless of the Fama-French exposures, i.e., whether using the phony SMB/HML=30%/27.5% or 22.5%/4.5%. The safe withdrawal rates in some of the pre-1929 banking crises were worse than those from a 75/25 portfolio. So, he’s still touring around telling people about his strategy, which looks so great, even endorsed by Big ERN’s simulation toolkit. But it’s not. Quite the opposite; the earlier retirement dates would have fared worse. I raised this issue at the 2025 Econome Conference when Frank presented his junk science numbers. Frank looked like a deer in the headlights in front of 200 people when I pointed that out. He’s still ignoring this caveat until today.

Of course, people may object that because the Fama-French factors didn’t start until 1926, we cannot simulate anything before that date. But that’s false. Because the SMB and HML factors are pure alpha factors, we can simulate before 1926 by setting the SMB and HML returns to zero while running the rest of the Risk Parity portfolio as usual. If Eugene Fama (a Nobel Laureate, by the way) and Ken French cannot construct the SMB and HML portfolios today, with all that data at their fingertips, then investors in 1911 couldn’t either. So it’s different from a scenario where one of the main portfolio components had no historical data. You couldn’t just set those returns to zero (say, pre-1871) and run simulations with zero total returns. But you certainly set the alpha factor to zero pre-1926. That’s 100% proper procedure.

Also note the great irony again: If Risk Parity is such a great strategy, it should also perform well on its own, without the help of the Fama-French alpha. By insisting on returns only post-1926, the Risk Parity + SCV crowd admits that Risk Parity on its own had been awful in historical cohorts where SCV alpha was unavailable. Dumpster Fries! This also aligns with the observation that Risk Parity + SCV has performed poorly again over the last 20 years compared with standard portfolios like 60/40 or 75/25.

10: The hodgepodge of 72 AI-generated complaints

Frank doesn’t have much to counter my two detailed blog posts; after all, he has one quantitative PhD, several quantitative and mathematical peer-reviewed academic journal publications, and one CFA charter less than I do. So, he can’t come up with anything intelligent himself. So, as a sign of desperation and the ultimate declaration of (intellectual) bankruptcy, he had the nerve to ask AI to conjure up a hodgepodge of crazy lies, strawman arguments, and baseless accusations. He placed the document in the show notes. Let me reply in more detail to a few of them:

“Rhetorical Belittling & Defamatory Framing” (21 items!)

So what? Someone who publishes such shoddy analysis that could jeopardize gullible listeners’ retirement security should be rebuffed in the starkest possible terms. He asks to be belittled because he has little education and little experience in asset allocation and portfolio optimization.

“Empirical Data Manipulation & Proxy Modeling: […] 5: Models DBMF using a static linear regression with a fixed -28% short equity exposure.”

That is a lie. And he goes on and on about his stupid lie in the podcast. AI complains that I have a fixed -28% short-equity drag in the DBMF simulation. AI doesn’t understand that the equity exposure in the DBMF simulation isn’t static, and it isn’t -28% either, because you have to add the ERN momentum strategy’s equity exposure (see Part 63 for details) with dynamic, not static(!) asset weights to get the actual dynamic net equity exposure. The net equity exposure in the DBMF replication varies between negative and positive because the ERN momentum strategy tactically allocates to equities; see the chart below with sample weights between 2000 and2025.

DBMF can short equities, while the ERN momentum strategy is long-only, so it’s no surprise that the static equity beta must be negative and the ERN (dynamic) momentum beta is positive. AI didn’t comprehend this, and someone as inept as Frank will find this offensive. But it’s exactly as expected and 100% intuitive for anyone familiar with such factor models (like yours truly, who’s actually taught econometrics at a university). Frank and his new best friend, the AI bot, are too dumb to grasp that.

Also, here is another great irony and logical contradiction. Look at the quote from Frank’s website, where he references a link to PortfolioCharts.com:

If you click that link, you will confirm that Tyler simulates the Golden Ratio strategy without managed-futures strategy returns. Instead, he uses a fixed/static(!), long-only commodity index (presumably the GSCI Total Return Index because it goes back to 1970). A quote from that link: “Frank specifically recommends managed futures, which is an asset I don’t have data for. I chose commodities as a stand-in”

Well, isn’t that interesting? Remember, Tyler is actually using a static-only asset allocation. And he’s ignoring other important asset classes (stocks and bonds) that trend-following strategies always use, while I clearly included stocks and bonds in my Momentum strategy (See Part 63). Tyler is committing the exact “data manipulation” that his AI falsely accused me of! Has Frank sent him an AI Scarlet Letter for such obvious simulation hackery? No, Frank is flattered. Frank and the AI slop have no issue with Tyler’s simulation, even though Tyler totally butchers the managed futures portion. Frank is not smart enough to notice these logical inconsistencies.

“and asymmetric fee friction is applied exclusively to alternative assets.”   and   “24: Modeled stock/bond benchmarks at 0.03% zero-friction costs, while baking a full 0.85% fee drag into the DBMF proxy.”

That is a lie. The opposite is the truth. Only an incredibly stupid AI agent would write that, and Frank, the “Chief Data Manipulator,” Vasquez has parroted this nonsense. My blog post clearly states that I don’t use the 0.85%; instead, I use the factor model (by the way, with a positive alpha, which translates into a negative(!) effective expense ratio for DBMF). The AI is too stupid to comprehend that. And Frank doesn’t notice it either. He should retract this AI nonsense!

“[G]old is evaluated using 100 years of government price-pegged data (1871–1971),”

Thanks! This supports my side! First, I don’t want to nitpick, but the peg era was March 1900 (Gold Standard Act of 1900) to August 1971. In any case, the gold-peg issue is exactly what I warn about when using gold exposure in retirement portfolio simulations. I point out this caveat every time I write about gold. Most importantly, I’m not the one using gold in my portfolio. Frank is, namely, 16% in the Golden Ratio portfolio, to be precise, which he blindly simulates through the price-pegged era! You are preaching to the choir, buddy! Frank is the one relying on suspect gold return data! Mark this one as a loss for Frank, generated by his own AI. You can’t make this up!

“requiring multi-asset models to survive high-CAPE regimes with 0% failure while giving traditional stock/bond baselines a pass for failing under identical shocks.” and “20: Applies a zero-failure standard under extreme overvaluation while tolerating 60/40 failures.”

That is a lie. I never gave 60/40 or 75/25 a pass. I always use consistency and comparability in my analysis. The failure and success criteria and the low/high initial CAPE conditional probabilities are calculated consistently across all portfolios. It’s a mystery why AI flagged this.

“16: Demands that all rules-based factor tilts be stripped out before a multi-asset portfolio is tested.” and “18: Systematically stripped out Small-Cap Value and Large Growth sub-units, replacing them with broad S&P 500.“

That is a lie. I don’t demand that. In fact, I obviously test the full Risk Parity portfolios with all bells and whistles, including SCV. I just included all the steps along the way, with and without SCV. I merely wanted to test different parts of the multi-asset portfolio to determine how much performance is due to each part. That sets apart careful, intelligent researchers from mathematically illiterate influencers. In finance and economics, we call it “attribution analysis.” That’s apparently too intelligent for this AI model to understand.

Also, my readers will acknowledge the great irony of Frank and his silly AI agent being so upset about me presenting the attribution analysis results where I (temporarily!) switch off the small-cap-value portion. They freely admit that SCV is such a crucial component in the Risk Parity simulations that the whole thing falls apart if you switch it off. When SCV is flat or slightly down (as it has been in the 2020s), the whole strategy folds like a lawn chair. My attribution analysis showed that the stellar performance in historical simulations has nothing to do with Risk Parity itself, but everything to do with historical Small-Cap Value alpha. And the angry reaction from the AI and Frank, accusing me of “Data Manipulation,” shows where the skeletons are hidden. Recall Section 6 above, where I showed that 75/25 with a Small-Cap Value bias would have outperformed Risk Parity! The same irony applies to returns pre-1926: Risk Parity is too pathetic on its own if you use zero SMB and HML alpha before the Fama-French database begins. Thus, Frank wants to avoid talking about the 1911 market peak at all costs! And Frank needs to desperately manipulate the simulations to 27.5% HML, even though his ETFs only have a net exposure of one-sixth that value (about 4.5%).

“14: Selected specific VIOV factor loadings ($SMB = 1.03$, $HML = 0.57$) to claim recent 20-year value factor drag. Ignores modern quality and profitability screening filters present in rules-based factor ETFs.”

That is a lie. VIOV and VUG are both completely passive with passive benchmarks (S&P Small 600 Value & S&P 500 Growth Index, respectively) and thus have no additional “modern quality and profitability screening filters.” They are the quintessential assets that this factor model can adequately replicate. VIOV and VUG had adjusted R^2 of 0.976 and 0.966, respectively. Even with additional filters (say, as in DFSV or AVUV, etc.), returns unexplained by beta would show up as alpha, which one can still replicate and simulate in the SWR toolkit, just as before. Besides, my AVUV factor model estimation had an adjusted R^2 of 0.965. DSFV’s was 0.967. The same ballpark as the purely passive ETFs. So, the active vs. passive discussion is a red herring anyway.

“Systematic, rules-based factor indexing is mischaracterized as active “stock picking,”

Semantics! Stock picking vs. Sector Picking vs. Style picking vs. any other strategy that deviates tactically or strategically from broad index weights is subject to uncertainty about what works when, why, and with how much tracking error. If anything, the prospect of gathering alpha is harder with SCV than with stock picking, exactly because SCV is purely rules-based. Stock pickers with some proprietary data source and/or data analysis tools can beat the market much more reliably than any rules-based style tilts today, when everybody knows the rules. In that case, Frank’s AI actually helps my case: by calling it stock picking, I gave the SCV style picking too much credit. Frank is also inconsistent with the previous point, where AI claims that I ignore active stock-picking flavors like “modern quality and profitability screening filters.”

“retail heuristics are dismissed for lacking 1,300-person institutional hedge fund staffs, “

A lie/strawman argument. I never claimed that. For the obtuse and reading-comprehension-challenged AI/Frank: Risk Parity alone is an inefficient asset allocation unless very stringent and unrealistic assumptions are met (the same Sharpe Ratio and all correlations are identical, which is clearly not the case). For a hedge fund to be successful, you need not just the risk parity concept but also plenty of alpha signals to generate competitive returns. I never claimed that retail investors can’t do Risk Parity without a staff of 1,300. They can’t do it profitably without research and trading staff. Without profitable alpha signals, Risk Parity investors have to resort to “Hail Mary” passes and cross their fingers that Fama-French alpha factors from 40+ years ago will miraculously start working again.

“11: Asserts modern investors shouldn’t use factor tilts because 1920s investors didn’t know about them.”

That’s a lie. I made no such assertion. I simply asserted that the alpha has disappeared, now that everyone knows about it. In finance, alpha strategies can persist when the wider public is oblivious. But once word gets out, the alpha shrinks or even disappears. Recall the Ed Thorpe anecdote above!

” 25: Evaluates multi-asset portfolios inside a rigid, non-flexible constant-dollar Monte Carlo simulation box.”

That is a lie. None of my Risk Parity work relies on Monte Carlo. It’s all complete historical cohorts. Besides, even if true, Frank uses my toolkit, so the same would apply to him. Let that sink in! Frank couldn’t run his own simulations if his life depended on it; he has to use my toolkit. Ironic, isn’t it? Why would he use the simulation results from someone he constantly accuses of data manipulation? He wants to have it both ways: nitpick my toolkit. But then still uses it because it’s the best simulation environment for the research questions at hand!

“26: Equates short-term portfolio drawdowns and trailing underperformance during bull runs to Sequence of Return.” and “59: Measures portfolio success by total return growth and keeping pace with 100% equities during bull markets.”

That is a lie. The horror scenarios in Frank’s Risk Parity portfolios are not mere underperformance in a bull market. All seven of Frank’s portfolios I studied that started before 12/31/2021 have had significant drawdowns (in real terms), and all face a huge SoRR headache. They are all below their starting values in CPI-adjusted dollars. That’s the definition of Sequence Risk. Your AI is too stupid to know and/or apply the SoRR definition and then accuses me of misdefining SoRR. Pure projection!

“31: Evaluates non-equity assets purely as underperforming substitutes for stock growth.” and “58: Embeds the premise that equities are the only true wealth engine into the claim that hedging is bad. Penalizes risk-hedged retirement portfolios for not matching S&P 500 gains in stock rallies.”

That’s a bunch of lies. I never claimed that. I’ve consistently acknowledged that including diversifying assets, especially T-bonds and T-bills, helps hedge Sequence Risk. Also, I report total return stats for reference only. I report performance for a retirement portfolio net of withdrawals, and Risk Parity performs even worse. Also, I never compare the Risk Parity portfolios to 100% S&P 500. I compared them to a 75/25 portfolio (and occasionally 60/40). See my table in the “How To Lie” post, Part 4, and the one in Section 1 above.

“34: Data Manipulation. Evaluated historical success strictly at an absolute $0 final asset target at Year 30, omitting median wealth distributions.”

Thanks! This supports my side! A Risk Parity portfolio, with its low equity weight, will have inferior outcomes when looking at the median final value. Thanks for pointing this out. You’re making my case stronger.

“61: Demands that non-equity assets exhibit immediate, guaranteed positive pops during every single stock drop.”

That’s a lie. I never made that demand. Quite the opposite; I’ve often lamented that in my favorite portfolio, 75/25, the bond duration doesn’t always hedge equity risk, e.g., in a supply-side shock and/or a Federal Reserve surprise rate hike. My 25% bond allocation took a beating during 2022, too. Nevertheless, I’m enjoying a rich retirement, while Frank needs to cut his retirement budget.

[V]olatile assets are evaluated on isolated trailing price charts rather than within a rebalanced portfolio system.” and “67: isolated standalone peak-to-trough asset drawdowns (e.g., -40% bonds, -50% commodities) to label them “toxic.” and “68: Evaluates standalone component drawdowns (-40% bonds, -50% GSCI) to declare the portfolio ruined.” Similar lies in items 64 and 65.

That’s a lie. I simulated complete portfolios with my toolkit, factoring in regular rebalancing. Only as an aside, and to add context and intuition, I plotted the total real return series for a few asset classes.

7: Frames grassroots, open-source educational discussions as commercial “false advertising.”

Bless your heart! False advertising isn’t limited to commercial transactions. Any distortion, obfuscation, deception, obscuring, concealing of facts, etc. – whatever word you want to use – in any environment can be called “False Advertising,” even in non-commercial settings. It would then be figurative, not literal. You’d have to be a really pedantic kindergartener to point this out. OK, after this one, I had enough. This is too stupid for me. Frank’s AI model is more artificial buffoonery. I’d like to know which engine he used, so I can all avoid that dumb drivel machine myself. Maybe it’s the free version of North Korea’s AI?!

So, you get the message; his list of 72 items is just a nothing-burger of lies and misunderstandings. Oops, that phrase “nothing-burger” is one the AI flagged in items 15+17, so I hope the AI doesn’t get stuck in an infinite loop when Frank sends this post through his Artificial “Intelligence” engine.

Many items are duplicates anyway (9+10, 16+17+18, 22+24, or 40+41, 64+65+67+68). Some are just nonsensical gibberish like “60: Applies a strict earnings-only filter (EMH) requiring underlying corporate dividend/earnings flows.”

AI also routinely contradicts itself: For example, AI complains that I incorrectly called SCV “stock picking” (#9,10). Then AI complains that when I simulate those systematic, passive, and rules-based ETFs with a factor model, I’m ignoring “the modern quality and profitability screening filters,” a.k.a. potentially stock-picking alpha in these funds (#14), which doesn’t exist in the VUG and VIOV, only maybe in the AVUV, DFSV, etc. Which one is it?

Another fallacy: the AI complains when I apply standard factor-model procedures and simulate the VIOV/VUG portfolio with estimated Fama-French betas (with remarkably high adjusted R^2, by the way). On the other hand, when Frank uses Fama-French betas in his sheet (and uses the wrong estimates, see item 7), the AI sees no issue.

So, this AI stunt has been a total embarrassment for Frank. He looks defeated and desperate. He’s lacking the experience and training to come up with his own reply. His AI just spat out useless slop. Most of it is false or irrelevant, and some of it actually supports my case. His AI output is even more useless than his usual buffoonish content, and that means something. Just in case Frank claims the AI is correct on all items I didn’t address in detail here, here’s my reply to all(!) 72 items.

Besides, I don’t want to dump too much on AI. It’s generally quite smart. To wit, I looked for references to where Frank made his unrealistic 5% SWR claim and searched with the prompt “bigger pockets frank vasquez guaranteed 5% withdrawal rate” to find the relevant BiggerPockets episodes. And guess what Google showed? My “How to Lie” post showed up prominently everywhere. It was the #1 reference in the list on the right, even above the BiggerPockets podcast pages I was searching for. And the warnings referenced my post as well. How cool is that?!

Lastly, I can play the AI game, too. For example, I asked both Google and Grok whether Karsten or Frank has more credibility in asset allocation, simulations, portfolio optimization, and finance. I included the responses in the document dealing with Frank’s AI Slop. Both agreed that I have more credibility. A quote from Grok: “Karsten Jeske (Big Ern of Early Retirement Now) has substantially higher credibility in asset allocation, quantitative simulations, portfolio optimization, and related finance topics.” Google had a great quote: “[Karsten’s] literal day job for a decade was creating, simulating, and optimizing institutional-grade portfolios for pension funds, endowments, and sovereign wealth funds”

Bonus: Frank agrees with me on something. Halleluja!

I saved the best for last: Frank gave me the best present possible in one of his prior podcasts: episode 546. In that podcast, he mentioned that he’s suddenly a big fan of the reverse glidepath in retirement. Well, isn’t that interesting? I have repeatedly pointed out that Risk Parity has insufficient equity exposure to sustain long retirements. However, 20- or even 30-year retirements can certainly survive lower equity percentages if you’re happy with asset depletion. At least we can agree on something! He won’t link to my research (Parts 19 and 20) on that topic, obviously (which I published in 2017, more than nine years ago, by the way), but who cares? But one of my readers who pointed this out was quite amused, and I got a good laugh out of it, too.

Conclusion

Frank operates in a carefully constructed system of obfuscation, data manipulation, and false advertising, among them:

• He is using vastly exaggerated Fama-French factor betas that he acquired from me under false pretenses (i.e., targeting a very different portfolio) to artificially boost his historical simulations, despite my repeated warnings. But he faults me when I’m using the proper Fama-French weights. Projection!
• Concealing his actual returns and only publishing simulations that benefited from hindsight and backfilling bias, i.e., replacing poorly performing ETFs that he had in his portfolio in real life with better alternatives in the simulations.
• On multiple podcasts, he misrepresented his portfolios as having drawdowns no longer than 3-4 years. However, his real-time performance post-2020 shows the opposite: stagnant growth and persistent drawdowns. All his portfolios have been in permanent drawdown status since the 1/1/2022 market peak. Now approaching five years of drawdowns!
• He also ignores the numerous drawdowns in the simulations (in the sheet that he so proudly uses), with drawdowns between 6 and 13+ years. And that’s before withdrawals. Drawdowns are even longer when factoring in withdrawals. Thus, the claim of a maximum 3-4-year drawdown is a brazen lie.
• Sneaky declines in the purchasing power of the portfolio withdrawals. When he appears on podcasts, he touts only the 5% initial rate, not the fact that his 5% withdrawals can become 4% along the way! His sample portfolios had to cut the withdrawals by between 5% and 58% over the last five years.
• He’s comparing his variable percentage withdrawal rates with the industry standard of fixed withdrawal rates. He’s comparing apples to oranges, because variable rates are naturally significantly higher.
• None of the portfolios I examined still withdraw 5% of their initial value, adjusted for CPI inflation.
• He is hiding the pre-1929 cohorts that would have fared far worse with Risk Parity.
• The Risk Parity portfolios are a bet on a revival of Small-Cap Value alpha. Without that alpha, Risk Parity would have had no edge or even worse outcomes in historical simulations. In the post-2020 era, with live returns, Risk Parity portfolios predictably underperformed when SCV alpha was nonexistent or even slightly negative.
• He has the nerve to post his shameful AI slop. It’s 72 items of either “Oh no, poor me, Karsten hurt my feelings,” or irrefutable lies, data manipulation, and strawman arguments.
• Selective outrage: he’s accusing me of data manipulation that I didn’t even commit. Then turns around and praises a post on PortfolioCharts that commits that exact data manipulation technique.

He does this knowingly, despite my repeated warnings about his faulty assumptions. He can’t claim ignorance; this is intentional! His Risk Parity claims fall apart under closer inspection. With the walls closing in and many of his sample portfolios already on a path to ruin (after only six years!), it’s clear he’s flailing and getting testy. To use a war analogy, I must be right above the target. I might drop some more “truth bombs” in the future if Frank wants to push the issue further.

Normally, I wouldn’t care. I can’t right every wrong on the internet. I can’t comment on every stupid hot take on Facebook and Twitter. If some random dude on the web blows up his portfolio, I don’t care. I start caring when Frank deceives his listeners with unsubstantiated claims about my research and my toolkit, using them to support his abstruse theories and inferior Risk Parity portfolios. They don’t. I have pointed this out to him multiple times, but he keeps pushing it, and this deserves the strongest possible pushback. I don’t want him to take my reputation down with him, and I don’t want his listeners to get the impression that my research endorses his Risk Parity portfolios in any way. 

Frank is a financial influencer and entertainer. And nothing more. He has no formal training, experience, or expertise in asset allocation and portfolio optimization. No accreditation. Even though he often references the CFA manual (probably 5 times in every podcast), I doubt he ever held one in his hands. I not only held the CFA materials; I studied and understood them, met all requirements, and earned my charter in 2010. In contrast to Frank, I’m in my wheelhouse here.

Frank’s advice endangers your retirement, as evidenced by the steep cuts in the withdrawal activity in his sample portfolios. His narcissistic “rarely correct but never in doubt” attitude extends beyond Risk Parity to other issues as well, like not understanding how CPI is constructed and falsely claiming that homeowners have only half the inflation rate in retirement; see Part 57 of my series.

Thus, I highly recommend taking all his advice with a grain of salt. Chances are, the exact opposite of Frank’s advice will be optimal. Get a second opinion before following any of his advice: consult a registered investment adviser. If you think my hourly rate is too expensive, consider how expensive Frank’s free advice could have been for gullible investors who followed him through the 2022 bear market with one of his Risk Parity portfolios!

I’ve got my popcorn ready and look forward to your comments and suggestions!

Picture Credit: WordPress AI