Investment Philosophy

Active vs Passive

Some investors adhere to what I believe is an elegant but outdated theory: the efficient markets hypothesis. According to this theory, all publicly available information is reflected in security prices. Assessing the quality of fundamentals, considering valuations, or any other activity used to pursue stock picking or security selection is futile. In other words, Warren Buffett was just lucky.

Some interpret the theory more loosely. They believe there are some opportunities, but the costs (fees, transaction costs, and tax friction) of exploiting them outweigh the benefits. Both of the above camps tend to pursue passive or index-based strategies whereby portfolios are constructed as slices of the overall market. This strategy effectively invests in all stocks, weighting each stock according to its market capitalization (i.e., more money is invested in bigger companies and less in smaller companies).

There are others, including ourselves, who believe markets are mostly efficient, but there are still security selection opportunities worth exploiting. Investors who attempt to take advantage of these opportunities using their own research are said to pursue active strategies. Active portfolio managers attempt to select individual securities to outperform their benchmark index (e.g., S&P 500). They tend to use a variety of techniques, including assessing fundamentals, valuations, price charts, and other analyses.

As Nobel laureate William Sharpe explained in his 1991 article, active management is a zero-sum game (over all time periods). For every active manager who outperformed, there must be one or more who underperformed. In other words, not everyone can be above average. Moreover, they likely charged higher fees and may have triggered taxes along the way. Standard & Poor’s reports performance figures of active managers in their S&P Index versus Active (SPIVA) scorecards, and the data clearly corroborates Sharpe’s arithmetic.

The active-passive debate has existed for years and continues to be ongoing. We feel the performance data regarding active funds is condemning. This makes a strong case for using passive/index-based strategies. At the same time, we also acknowledge that there are portfolio managers who have consistently outperformed market benchmarks, and we believe some are likely to continue to do so.

To be sure, historical outperformance is not synonymous with future outperformance. Accordingly, we leverage tools that we developed to help us attribute historical outperformance and assess whether it was statistically significant (i.e., was it luck?). This further screening is critical, as the active fund industry is skilled at hiding underperformance (see this study or an article discussing it).

On balance, we generally use index-based funds for the bulk of our portfolios. However, we also prefer to use specific active funds where we see specific benefits in the context of outperformance or diversification.

Note: Aaron teaches many of the concepts above in his Equity Portfolio Strategies class for the University of Florida’s Master’s in Finance program.

ETFs & Mutual Funds

Many people conflate the notions of index funds and exchange-traded funds (ETFs). The term “index” refers to the (passive) investment strategy pursued by the funds managers. However, the term “ETF” refers to the investment vehicle or legal entity in which the investments are managed.

One of the primary benefits of many ETFs is their low cost. Indeed, fees on many ETFs (and index funds) are measured as small fractions of a percent. For example, the Vanguard Dividend Appreciation Index Fund ETF (ticker: VIG) boasts an expense ratio of just 0.05%.

There are even funds with 0.0% expense ratios! There are some publicity benefits, but it is worth noting that expense ratio investor pay are not the only source of revenue for fund companies; these funds regularly earn additional fees by lending shares held within the fund to other entities (e.g., hedge funds or banks who want to short the stock).

Note: When using index funds, it is important to understand the mechanics of the indices underlying the funds. In particular, the rules for how an index selects and weights its holdings ultimately determine how the index and index fund will perform. While the benefits of index funds are well-known (e.g., low costs and tax efficiency), few are aware of the potential drawbacks.

In addition to the low costs, people often appreciate the tax efficiency of ETFs. While this tax efficiency is commonly attributed to the low-turnover nature of the portfolios, ETF tax advantages are deeper than that. Indeed, ETFs enjoy a tax blessing from the IRS.

This tax advantage primarily stems from their “in-kind” creation and redemption mechanism. It effectively allows them to mitigate, if not avoid, capital gains distributions. This creates a significant advantage relative to mutual funds, and is why many active mutual funds are converting to ETFs.

We discuss this further on our Tax Strategies page and highlight more details in our video on ETFs, but we also highlight a classic case study below.

The Sequoia Fund: A Case Study for the Taxation of Mutual Funds

In 2015, the Sequoia Fund had a heavy concentration in one stock, Valeant Pharmaceuticals (now Bausch Health Companies Inc). After a series of questionable business and accounting practices came to light, Valeant’s stock price plunged over 95%, including a 60% decline within a single week.

Many investors in the Sequoia Fund were disappointed and redeemed their holdings in the fund. The fund made efforts to limit taxable gains by taking losses in Valeant and making in-kind redemptions (i.e., paying investors with shares of stock the fund owned, rather than cash). Ultimately, there were too many redemptions, and they also had to sell highly appreciated holdings.

On balance, the redemptions for exiting investors resulted in significant capital gains for investors who remained with the fund. Despite experiencing losses of almost 7% in 2016 (while the S&P 500 was up nearly 12%), the loyal investors who stuck around also received tax bills. This salt in the wound is a poignant example of how investors are not in control of their taxes with mutual funds.

As a financial planner who optimizes taxes around various brackets and thresholds, we prefer to maintain as much control of taxable gains and thus try to avoid using mutual funds in taxable accounts.

Inflation

Bill Bengen, the originator of the 4% withdrawal rule, says that inflation is the number one enemy of retirees. Indeed, unexpected increases in inflation typically hurt both stock and bond portfolios. At the same time, it increases the level of spending required to maintain the same standard of living. In other words, inflation can simultaneously reduce the assets that will provide for retirement while increasing the liabilities.

In periods where inflation has been significantly higher than average, this double-whammy blow has historically crippled retirees and significantly reduced sustainable withdrawal rates.

Some tools to combat inflation:

• Social Security (SS): Your SS benefits are formulaically linked to inflation. Even before you start taking benefits, the SSA adjusts your primary insurance amount or PIA (i.e., the figure on which your future benefits will be based) according to the Consumer Price Index for Urban Wage Earners and Clerical Workers (CPI-W). These inflation adjustments continue when you start taking your benefits, too. This is another reason why delaying benefits is often a good idea; higher SS payments translate into higher inflation adjustments, which can help protect you’re retirement against this risk (inflation).
• Stocks and dividends: Stocks get hit in the short term when inflation strikes. However, they actually have positive correlations with inflation over longer periods, as companies strive to maintain their margins and pass inflation through to their customers. We have written research that describes this phenomenon and backs it  up with data (please see Dividends: Theory and Empirical Evidence). Historically, this has presented a tremendous opportunity where investors could turn lemons into lemonade.
• Commodities: Commodities naturally perform well during inflationary periods. However, they can be dead weights when inflation is stable or low. Given that we do not believe we have the ability to predict when inflation will rear its ugly head, we do not like holding commodities.
• Alternative investments: Some types of alternative investments may be helpful for combating inflation risk. Please see the Diversification section where we discuss this in more detail.

It is also worth noting that portfolios typically require a minimum level of stocks to ensure that we have sufficient inflation protection over the long term.

Dividends

Put simply, we believe dividends represent the unique mechanism by which companies can return cash to shareholders without market interference. Moreover, dividends have historically proven to be remarkably robust and correlate strongly with inflation. We believe these attributes make dividends particularly useful in the context of generating retirement income.

Many people claim that synthetic dividends (i.e., chiseling off principal from stock holdings) are economically equivalent to traditional dividends paid by the corporations. While this is almost universally accepted as conventional wisdom, it simply is not true. We explain this in more detail and provide empirical evidence within the articles we highlight below, but here are two simple notions we use to defend our position:

1. Dividends are fundamental in nature, as they are paid directly from corporate balance sheets to investors. Indeed, these are corporate decisions and do not depend on whether stock prices are up or down that day. They also do not depend on whether the overall market is sitting at low valuations after a sell-off (e.g., right after the pandemic surfaced) or record high valuations after years of optimism and growth.
2. If one is generating their own synthetic dividends by chiseling from one or more stock positions, they are naturally at the mercy of what the market offers when they sell some of their shares. If the market is up or down that day, they will have to sell fewer or more shares to generate the same amount of money. On any given day, the market is bombarded with seemingly random news that could push it up or down (presidential Tweets!?). In other words, market noise impacts the economics. 

Given that many investors and investment professionals acknowledge the risk around the sequence of returns, I am surprised that more people do not recognize the role that dividends can play as a more stable and market-independent source of income.

To be sure, we also acknowledge that dividends are not a holy grail, and they have some issues. For example, we do not advocate pursuing high-yield dividend strategies (e.g., see the Tax Strategies page where we highlight related risks and tax issues in the Core Strategies section).

For those who are interested in learning more about this topic, we have written two long-form research notes on the topic:

• Dividends Are Different
• Dividends: Theory and Empirical Evidence

Diversification

Diversification is a well-known concept, but it is essential to make sure that the assets one owns provide genuine diversification benefits. In many cases, we see portfolios that were seemingly built with a “dash of this and a dash of that” mentality. Even worse, we regularly see portfolios where the funds and portfolio managers were chosen because they kick back additional fees to the advisor (see this Barron’s article describing these pay-to-play revenue-sharing arrangements).

We believe it is important to build portfolios purposefully. That is, one should select each ingredient based on how we anticipate it will perform, not just on its own (i.e., in a vacuum), but also how it will behave relative to other investments within the broader portfolio.

Here are three examples of how we think about diversification:

(1) “Humpty Dumpty” portfolios: We often see portfolios with many different stocks and/or stock fund holdings. Unfortunately, more often than not, the strategies from one fund counter what other funds are doing. So, when we look at the portfolio in aggregate, it looks almost identical to the overall market. For example, one of the most common portfolio combinations includes small, medium and large cap stock funds. These types of funds basically break the overall market up into three pieces. Thus, when someone tapes them back together within the same portfolio, their combined performance is often indistinguishable from a total stock market fund.

(2) Alternative asset allocation: Another area where diversification is critical is at the asset allocation level. There is much buzz around private and alternative investments, and how they can reduce volatility and improve diversification. We have two thoughts on this.

First, just because one cannot see daily prices with private investments, it does not mean they are not volatile. Second, most of these investments are linked to the underlying economy. In particular, they may be different types of investments, but they can exhibit strong correlations with other risky assets (i.e., stocks).

Real estate is a perfect example. Not only is it linked to the overall economy, but a significant amount of real estate has been securitized. That means it is likely susceptible to the emotions and whims of millions of investors, just like the stock market. As a result, we expect real estate to have higher correlations with the overall stock market (i.e., less diversification benefits).

We suspect much of the attention given to these investments is primarily driven by the brokers and banks that earn commissions from selling them. Notwithstanding, we believe there are some sub-classes of alternative assets that provide valuable diversification.

In particular, our fundamental modeling indicates that some alternative investment strategies should be able to help mitigate risks related to inflation. This is especially important, as inflation is likely the most important risk retirees face (please be sure to see the section on inflation). While one year of performance is only anecdotal, we were pleased to see these true diversification benefits play out when inflation reared its ugly head in 2022.

(3) Corporate and high-yield bonds: The value of corporate bonds naturally depends on the health of the underlying companies. If a company is doing well and generating strong cash flows, the market generally anticipates little risk and imposes a smaller discount (i.e., credit spread) relative to comparable risk-free treasury bonds. However, if a company experiences distress or heads toward bankruptcy, both its stock and bonds will likely decline.

Academic models dating back to the 1970s have highlighted this connection between corporate bonds and stocks. Moreover, historical price correlations provide additional empirical evidence supporting the theory. While this is just one example, we believe it is worth noting that the corporate bond market declined in tandem with stocks when the coronavirus pandemic first emerged. In other words, corporate bonds provided little, if any, diversification benefits during that period.

When diversifying a portfolio, we believe it is essential to utilize ingredients that have low or negative correlations. This becomes especially critical during volatile periods. This is why we tend to make smaller allocations to corporate bonds within our fixed income portfolios.

(3) Real estate: Many investors and investment professionals utilize real estate as an asset class. Moreover, their allocations on par with those made to stocks and bonds. We utilize smaller allocations to real estate for two primary reasons.

The first reason is that investors in the stock market already own a significant amount of real estate through the companies they own. The second reason is that real estate is cyclical in nature and highly correlation with stocks. Moreover, real estate has increasingly been securitized (e.g., real estate investment trusts, or REITs).

As more people own real estate through liquid investment vehicles, it becomes more exposed to the whims and sentiments of millions of investors. On balance, we find that its role as a diversifier has diminished.

As we highlight in the Data & Modeling tab, we live in a data-rich world, and hordes of CFAs enjoy applying elegant mathematical models to historical market price data. However, significant issues arise when analyzing risk in this way. 

For starters, the old adage “past performance is not necessarily indicative of future performance” is naturally relevant. While the topic is quite technical, my experience has shown that correlations are an Achilles’ heel in financial modeling.

In order to run simulations, one needs to integrate how different assets perform relative to each other. The most common approaches involve the use of correlation or covariance matrices (tables of pairwise correlations/covariance across assets or asset classes).

While correlation may seem straightforward or benign in modeling, it is a nuanced concept. Interestingly, one of the focal points during my previous career in equity derivatives was correlation. By comparing the level of volatility embedded in index options (e.g., calls and puts) to the levels of volatility embedded across the options on individual stocks, we could impute what level of correlation was baked into the options market.

One issue that arises is that correlation assumes relationships are symmetric in nature. However, many relationships in the financial markets follow more of a cause-and-effect relationship.  When asset A moves in one direction, it might be more likely that asset B also moves that way. However, the converse may not hold true. That is, aasset B does not cause asset A to move, or at least not as much as A does to B.

We believe it is important to understand what is going on beneath the surface of market prices. In particular, we assess how the fundamentals of some assets perform relative to the fundamentals of other assets. Knowing that markets eventually follow fundamentals, we find that this approach to modeling and diversification can be more effective. Indeed, if can help strip out the noise imposed by market participants that plagues analyses based on market price data.

Asset Allocation

Asset allocation refers to how much money we allocate to different asset classes (stocks, bonds, real estate, etc.). It is typically the most significant factor impacting how portfolios perform. In general, stocks have higher expected returns over the long run, but also more volatility along the way. That is why we make the stock allocation the primary focal point of our asset allocation process.

Note: While our asset allocation process focuses on the liquid portfolio, we take into account other assets and sources of cash flows (real estate, SS, pensions, annuities, etc.).

When determining what asset allocations are sensible for our clients, we assess three key dimensions. The first dimension is risk capacity and relates to how much risk one can afford to take. This element is purely by the numbers. We use simulations to investigate what levels of risk (allocation to stocks) are likely to deliver on clients’ monetary goals.

This generally translates into a range of allocations. The lower end of the range represents the minimum allocation that will still provide sufficient protection against inflation (see the inflation tab). The upper end of the range represents the level at which higher allocations to stocks might open the door to too much volatility and jeopardize financial security.

The second dimension of our asset allocation assessment focuses on risk tolerance. That is, how much risk can one stomach? Even if a portfolio strategy is likely to help clients meet their financial goals, they must be sufficiently comfortable with the strategy, especially during more turbulent periods.

Of course, nobody enjoys seeing the value of their portfolio decline. However, one of the worst situations can occur when one discovers they cannot tolerate their portfolio after it has declined.

Our disciplined portfolio process generally takes advantage of market declines by rebalancing and allocating more money to stocks after significant declines. Indeed, our goal is to restore and maintain our target asset allocation. However, volatility can trigger investor emotions that make them want to go in the opposite direction.

While not intentional, emotions can lead to a buy-high/sell-low phenomenon that can significantly hamper long-term performance. This makes it critical to ensure we build portfolios that our clients can stick with through thick and thin.

This aspect of risk management also emphasizes the value of having an investment advisor. Not only will they remain objective and pursue a disciplined portfolio strategy, but they can also reassure clients that their financial plans account for and are built to withstand volatility without compromising their long-term goals.

Lastly, the third dimension of risk is what we call risk sensibility. Just because one can afford and tolerate risk-taking does not make it sensible. For example, there may be periods when the valuations of stocks, bonds, or other assets are on the high (low) side of history. In these situations, we prefer to lean toward the lower (higher) end of the allocation ranges determined by one’s risk capacity and tolerance.

While defining sensibility in this way may overlap with the notion of market timing, the historical correlations between starting valuations and long-term investment performance have been strong (80-90% and higher). Economic theory and intuition also support this notion that buying something at a higher price naturally reduces its future returns.

Given that we live in a world where stock prices and news are constantly available, it can be harder for us to take a step back and see the broader trends. However, we view this task as part of our responsibility as financial planners; we highlight factors that can impact our clients’ planning.

At the same time, the usual disclaimer applies: “Past performance is not necessarily indicative of future performance.” We do not have a crystal ball and acknowledge that the economic theory and historical factors we observe may have less relevance in the future.

The bottom line is that asset allocation is a mix of art and science. I would argue there is more science than art. So, it is important to pursue a disciplined process. This includes ensuring that portfolios are compatible with clients’ goals and risk tolerances. However, we should also remain humble and not overestimate the precision of our models, simulations, and analyses.

Structured Income

One natural extension of my quality and value approach is the construction of portfolios targeting steady and increasing income. Despite a seemingly insatiable demand for stable income, I find the existing market of funds and other investment products falls short in delivering on this demand. Indeed, many funds (even those whose mandates mention dividends and income) typically generate very unstable income streams.

Our research highlights this income instability issue and explains why it exists. For example, I find the top priority for portfolio managers is the pursuit of total return – the metric by which virtually every investor, investment database, analysis, and regulator ultimately judges them. Unfortunately, focusing on total return compromises the income stream investors ultimately experience. Indeed, income becomes an ancillary factor or is completely ignored when securities are purchased and sold.

Given the lack of stable income generated by existing investment products, it is no surprise that many investors become interested in annuities and other income-focused products. While these products are not all bad, it is important to understand their mechanics and pricing. It is always important to work out the math and figure out what returns and fees are embedded in these products.

Please read this article or watch this video to learn how I structure steady and increasing income streams while preserving and growing your wealth. This could be one of the most refreshing approaches to retirement income you ever encounter.

Annuities: The Ugly, the Bad, and the Good.

There are many negative stigmas attached to annuities, and for good reason. The commission-based nature of this business incentivizes agents to sell these products regardless of whether or not they are a good fit for a client. In particular, the regulation of annuity sales only requires that they have to be “suitable” for consumers.

We cannot understand how much this differs from a fiduciary standard of care where one must do that they believe is “best” for the client. We have observed numerous cases where consumers were sold annuities based on deceptive marketing pitches. This often left the annuity owners with a combination of high fees and lackluster performance while they were stuck with the product, as well as steep surrender penalties to get out.

We also find that many of these products are designed primarily for marketing purposes, rather than optimizing outcomes for consumers. For example, many of the index annuities offer upfront bonuses and guaranteed crediting. However, these benefits are ultimately diluted by a product’s payout ratio (i.e., the rate at which it converts values based on those other benefits to guaranteed lifetime income). In our experience, these payout ratios are far below the rates at which simpler annuities could convert assets into income.

Notwithstanding the potentially negative aspects of annuities, we do not believe they are all bad. Indeed, we find annuities helpful in many cases. While annuities are not a part of all of our financial plans, we find they can be efficient vehicles for augmenting other sources of lifetime income (e.g., SS and pensions).

If an annuity make sense and a clients is interested, we can quote dozens of insurance carriers and product lines (e.g., fixed and indexed). Given the numbers of carriers and the slow pace at which many update their annuity pricing, we regularly encounter attractive opportunities.

With our actuarial annuity calculator and financial planning tools, we can make sure the economics make sense and are compatible with one’s broader plan for income and tax efficiency.

Life Insurance

We believe the the old adage that “life insurance is sold, not bought” is still very relevant. Indeed, many agents will push life insurance products even when they are not a good fit. At the same time, I also believe there are situations where life insurance can be appropriate and fit specific needs.

As most people near or enter retirement, there is less need for life insurance as a tool for income replacement. However, it can also serve as a tax planning tool. For example, the many married couples are exposed to risks stemming from what is known as the widow’s tax.

After one spouse passes, the surviving spouse will be filing taxes individually. In particular, the thresholds for tax brackets will compress and their tax bills may go up even if they have less income (e.g., lose a Social Security benefit). This phenomenon is knows as the widow’s tax. This risk is especially relevant in situations where couples have a significant amount of their assets in pre-tax accounts (e.g., IRAs).

Life insurance can also serve as a modular and tax-efficient tool to leave money to heirs. However, the bottom line is that life insurance may still be relevant during retirement.

If and when life insurance may be relevant, we use our actuarial life insurance calculator and financial planning tools to ensure the economics make sense within the context of one’s broader financial plan.

Other Products

Disclaimer: The following discusses investment products that directly or indirectly use options. As such, this content falls on the more technical side of the spectrum.

Some Background

People regularly ask us what we think of various investment products. One of the more popular flavors of the day is the “defined outcome” fund. These funds typically embed a stock market index as their core position, but then overlay that position with options (calls and puts).

For example, these funds will typically purchase put options to protect against major market downturns. At the same time, they will often sell call options to help fund the put purchases. These short call positions naturally limit the upside performance as well.

Option overlay strategies such as these are effectively risk management solutions. They still allow some upside and some downside, but the performance is limited in both directions. While these puts and calls are typically struck out-of-the-money, it might be intuive to think about two alternative versions of this strategy:

1. First, let’s consider an example whereby we push the strikes far out-of-the money. That leaves plenty of room for the funds to appreciate or decline. This strategy would perform closer to a raw position in the underlying index.
2. Second, let’s consider an example whereby we pull the strikes inward so that they are at-the-money options. In this case, we have removed all of the upside and downside. Consequently, we have no stock market risk, and should thus earn the risk-free rate.

Before we offer our perspective on these products, let us first point out that these strategies have been around for many years, even decades. For example, banks have and continue to issue structured products with various option overlays (enough combinations of calls and puts to make your head spin). Insurance companies also got in on the game with their index-based accumulation annuities.

We will brush the technical distinctions between these products aside, as we suspect nobody reading this is likely to know about (or want to know about) core differentiating factors such as cliquet options. Suffice it to say, options strategies have been around for a long time, but are now being offered through ETFs, which makes them easier (and presumably cheaper) to access.

Our Perspective

While many of these investment products are very popular, we do not care for them. There are three primary reasons we do not use these products. The first is that the cost of directly hedging risk is typically the most expensive way to do it. In the world of options, (out-of-the-money) puts are relatively more expensive, and this is reflected in something called the volatility skew.

The products and risks we are discussing relate to the stock market, but the cost of hedging is generally expensive. Put another way, risk is like a hot potato. That is, you generally have to pay someone extra (profit) to warehouse risk.

The second reason we dislike option-based products is that they increase path dependence. That is, the payoffs of the options will alter the overall performance based on the trajectory of the market through time. For example, the performance of option-based strategies will differ if the market goes up and then down, down and then up, or just goes sideways. However, the performance is the same with a buy-and-hold strategy.

In the context of retirement (our primary focus) and withdrawing from portfolios, this exacerbates the already-known issue of sequence-of-return risk.

The third reason, which overlaps with the second, is that the constrained payouts with option overlays likely weakens the fundamental relationship between inflation and performance. As we discuss in the section on inflation, a core reason we maintain significant allocations to stocks is that they correlate with inflation over longer time periods. However, by manipulating market performance and making it more path dependent with options strategies, we are likely interfering with this fundamental signal.

Data & Modeling

We find that academics and others analyze market price data because it is so plentiful. Moreover, it allows them to leverage elegant (but not always practical!) mathematical and financial models.

When it comes to data, the most plentiful type is market price data. We can find historical market prices for most investments freely available on the internet. However, it is important to understand that market prices embed a significant amount of noise.

While we believe prices ultimately follow the fundamentals, they can also swing far away from the fundamentals for extended periods. This noise can open the door to garbage-in-garbage-out (GIGO) issues when market data is used to look into the future.

“In the short run, the market is a voting machine but in the long run, it is a weighing machine.” – Benjamin Graham (Warren Buffet’s mentor)

Data aside, it is also important to understand the limits of financial models. This includes many of the formulas, simulations, and rules of thumb that are widely quoted and used. Many of these concepts were developed within academia, but have significant limitations in the real world. Some examples include:

• Volatility: This quantity is typically calculated as the standard deviation of historical investment returns. Many popular financial planning software packages use this figure to represent risk. However, I have yet to meet a client who feels that the market gyrations embodied in this calculation represent the risk that impacts their psychology.
• Simulations: Monte Carlo simulations (MCS) are very popular with financial planners (us included). However, it is important to understand the limitations of MCS, especially in the context of retirement planning. For example, correlations are an infamously difficult challenge in the world of MCS. They do not properly reflect the underlying or structural relationships between asset classes, they are linear in nature (i.e., not compatible with tail risk modeling – an important concept for retirement!), and there are technical issues related to how to replicate correlations within MCS (see this Wikipedia article). Moreover, the memory-less stochastic variables used within MCS do not properly capture the empirical mean reversion we see in the empirical data (see this article from Michael Kitces’ blog).

“All models are wrong, but some are useful.” – George E. P. Box (British statistician)

On balance, we find historical performance is helpful for informing investment decisions. At the same time, we acknowledge the limits of our data and modeling.