✨ How To Create A Low-Volatility Value Investing Portfolio
✨ How To Create A Low-Volatility Value Investing Portfolio
Want to invest The Warren Buffett Way, but don't have the stomach for volatility? Click here for the solution!
Value investors tend to be volatility-agnostic - which might not be a palatable investment approach to many.
By pairing Factor Investing with Value Investing, it is possible to create a Low-Volatility Value Investing Portfolio.
This has the benefit of appealing to both crowds - albeit involving some compromises. However, the aggregate value-add can be phenomenal for all parties involved.
Some of the benefits of practicing a Low-Volatility Value Investing Portfolio include - delivering both consistently superior portfolio risk:reward and low volatility; attracting a larger AUM base (e.g. institutional investors); and making The Warren Buffett Way more accessible to investors concerned with short-term volatility.
One of the most common questions I get when I reveal that I self-identify as a value investor is - how do you manage Volatility?
Since value investors like Buffett and Munger tend to invest in companies purely based on the merit of their underlying fundamentals - rather than the possible trajectory of their share prices - little consideration tends to be given to the volatility of their portfolios in investment decisions. Hence, their investment portfolios tend to exhibit outsized volatility as compared to those with non-value investing mandates.
One reason for this preference amongst value investors to ignore volatility in their investment decisions is because we do not consider Volatility As Risk (VAR) - which many more intelligent people than me have gone to great lengths to explain why. Personally, I’ve also gone on a mini-rant in my article below about the shortcomings of the CAPM model as a measure of cost of equity:
Allow me to utilize this paragraph to give you the abridged version. The volatility of share prices is encapsulated in the risk measure of beta - which basically involves a regression analysis of historical share prices in an attempt to figure out their aggregate deviation from the mean trendline. Here’s the problem: If beta indeed represents risk, how does the volatility of historical share prices inform an investor of the potential future risk of a fire breaking out in a particular business’s warehouse? This could only be remotely true if historical share prices somehow contained perfect information - wherein extrapolating the historical mean trendline of share prices could potentially yield credible information about the future mean trendline of share prices; and where a deviation from that extrapolated future trendline could potentially inform the risk of that extrapolated trendline being unreliable for forecasting purposes.
As Munger likes to call it, that’s all “hogwash”. Share prices do not behave according to neat statistical rules in real-life; and risk in the real world is almost entirely unreflected in the beta. Don’t take my word for it; take Warren Buffett’s.
Having said that, the inescapable reality is that the wider investment industry isn’t exclusively made up of value investors - and that there are millions upon millions of industry practitioners out there who still subscribe to this financial fallacy. One way to possibly justify the use of Volatility As Risk (VAR) is that the vast majority of investors in open-ended funds tend to be short-term focused - and that excessive portfolio volatility does indeed contribute significantly to investor redemptions, owing to the perception of risk. And if an investor redeems their funds from the portfolio in the “short-term”, quite naturally the “long-term” doesn’t even get a chance to manifest.
Regardless of whether the theoretical debate of VAR is justified or not, the fact remains that this is the main reason why the investment industry prioritizes volatility as a measure of risk - oftentimes even at the expense of actual investment risk. As the recent cratering of Tech sector valuations have amply demonstrated, investors who haven’t given sufficient consideration to the possibility of their analysis being wrong tend to get shaken out of their positions when share prices crater - since to the unintelligent investor, volatility in the absence of risk is only visible in hindsight.
As a result, many professional investors - in their endeavor to minimize relative portfolio volatility (i.e. beta) - tend to become closet indexers or own satellite portfolios, where the bulk of their portfolios mimic the passive composition of their benchmark index with only a small portfolio weight allocated towards an active “alpha” component. Of course, the tradeoff involved here is lower relative outperformance for lower relative volatility - since if the majority of your portfolio mimics the benchmark index components, the chance of relative outperformance to the index is also reduced.
I have many, many thoughts about the shortcomings of such an approach (both philosophically and practically) - but it would take me an entirely separate essay to rant about them, hence I shall refrain. However, I do intend to provide a solution to such shortcomings in this article - which allows you to leapfrog them in a practical investment context, regardless of whether you agree with the philosophical narrative of Volatility as Risk or not.
In essence, the problem that we are attempting to solve today is this - if minimizing volatility is an unavoidable reality in professional settings, how do we continue to practice value investing in a professional setting while still attaining a low volatility portfolio?
There are several benefits to doing this. Firstly, it would obviously reduce portfolio volatility to acceptable levels - making such a strategy palatable to the vast majority of market participants who tend to prioritize short-term volatility over long-term outperformance.
Secondly, it would enable a higher probability of portfolio outperformance relative to the benchmark index. This can be achieved simply by virtue of having a different composition of stocks - but also because we are actually optimizing for portfolio risk:reward in real-life; as opposed to observing some statistical measure of risk with no tangible root to reality. In layman’s terms, you are giving the portfolio more room for potential upside while at the same time narrowing your room for potential downside - without fooling yourself into thinking that your actual downside has decreased just because some statistical formula says so.
Hence at the risk of repeating myself, the question we are trying to solve today is - How Do We Create a Low-Volatility Value Investing Portfolio? Let’s find out.
Understanding the Factors which contribute to Portfolio Volatility
In financial speak, fund managers are able to do something called an attribution analysis to identify which ‘part’ of the portfolio has contributed to which outcome at the aggregate portfolio level. The official industry terminology for these ‘parts’ is called “factors” - e.g. the “Value” factor or the “Growth” factor. So if the Tech sector (which has historically sported high earnings multiples) underperformed in the past quarter, we might “attribute” the bulk of quarterly underperformance at the aggregate portfolio level to the “Growth” factor.
This factor attribution isn’t merely constrained to “style” factors - they can literally represent anything. You might carve out the Tech sector as its own factor and attribute a portion of portfolio outperformance to it - just as easily as you might do the same for the Inflation factor, or even the Asia ex-Japan factor. Basically, this is just an arbitrary ex-post attribution of portfolio performance to a certain part of the portfolio - in an attempt to identify segmental trends with respect to portfolio performance.
While relying on attribution analysis for actionable insight into future portfolio performance is questionable at best, this is not to say that there is zero value in doing it. There is an indubitable truth in certain factors having enduring correlations with each other - e.g. Equities might have an inverse correlation with Fixed Income, while credit impulse tends to have a positive correlation with GDP. This is the heart of attribution analysis - identifying what factors have correlations with one another, and then aggregating the respective factors to achieve a target portfolio with the desired ex-ante portfolio characteristics.
Allow me provide an oversimplified example. Suppose you were worried about the impact of future interest rate trajectory on your portfolio - and wanted to neutralize interest rate risk as a portfolio objective. An easy way to neutralize interest rate risk would be to buy variable fixed income instruments - so that whichever way policy rates move, the interest rate of your variable fixed income securities would also follow, thus negating interest rate risk no matter which way policy rates swing (again, this is oversimplified).
Another more complicated way of neutralizing interest rate risk - without involving the tradeoffs in allocating portfolio capital towards variable fixed income securities - would be to pair Financial sector stocks with Real Estate stocks (once again, this is a gross oversimplification). As companies in the Financial sector tend to experience wider lending spreads when interest rates rise, their share prices would tend to outperform as the latter rises. Conversely, as many consumers of real estate products tend to rely on bank financing for their purchases, higher interest rates tend to dampen demand for real estate products and thus reduce activity in the Real Estate sector - resulting in the share prices of these companies underperforming as interest rates rise. As such, we might be able to say that the Financial sector has an inverse correlation with the Real Estate sector - purely in the context of interest rates.
Thus, if we wanted to neutralize interest rate risk at the aggregate portfolio level (in this oversimplified scenario), we could in theory allocate portfolio capital evenly between the Financial sector and the Real Estate sector - resulting in the aggregate portfolio experiencing a neutral impact to interest rates, regardless of which way they turn. Subsequently, we could also have confidence that interest rates would have zero attribution to our portfolio’s performance going forward, as long as it remained evenly allocated between these two sectors - which might have utility towards reducing portfolio volatility in these highly uncertain times with respect to policy rates.
Similarly, we can also do the same for all relevant portfolio ‘factors’ in order to achieve a targeted ex-ante portfolio exposure. Let’s say that the index which your portfolio is benchmarked against is the MSCI World Index. You could in theory perform an attribution analysis on the MSCI World Index to figure out what the factor composition of the MSCI World Index was - and then construct a portfolio with a similar factor exposure. In the same way that allocating your portfolio evenly between the Financial and Real Estate sectors in the example above neutralized interest rate risk, attempting to mimic the factor exposure of the MSCI World Index could in theory allow us to minimize portfolio volatility relative to the benchmark index (i.e. attain a portfolio beta of 1.0).
The beauty of this portfolio strategy - in order to reduce portfolio volatility - starts being apparent when compared against the aforementioned method of becoming a closet indexer (e.g. satellite portfolio). When using the latter approach, one is only able to reduce relative portfolio volatility at the expense of portfolio outperformance - since your portfolio is largely composed of the same stocks as the benchmark index components. However, when using the factor approach as described above, your portfolio’s outperformance isn’t necessarily constrained by the need to invest in the same stocks as the index - since each individual stock can have several different factors, and which individual stock position you hold matters less than the aggregate factor exposure of the overall portfolio. In theory, you could use attribution analysis to figure out which stock possessed which factor - and then mix-and-match stocks according to your liking to achieve your targeted ex-ante portfolio exposure.
If the portfolio objective was to minimize portfolio volatility relative to the benchmark index, the targeted ex-ante portfolio exposure would be attaining a portfolio beta of 1.0. Hence, how the aforementioned portfolio strategy would play out is that the fund manager would simply mix-and-match different stocks in an iterative fashion until he arrived at an ex-ante portfolio beta of 1.0 - and iterate the process everytime he wanted to rebalance the portfolio or add/remove a position. In this manner, the fund manager is able to consistently minimize relative portfolio volatility.
This is a process which can be easily automated - indeed, there are plenty of OTC portfolio management products such as the MSCI Barra Optimizer, which can automate this process for you. Or you could simply hire a junior analyst to iterate the process using Microsoft Excel - I know this because I was once tasked with this exact task before.
Where Value Investing Comes In
At this point, we have managed to achieve a low-volatility portfolio by applying factor attribution to portfolio management (i.e. lower volatility). However, this still begs the question - how do we achieve portfolio outperformance with this method (i.e. higher return)?
To recap, everything we’ve described above only lends itself towards the objective of attaining a targeted portfolio volatility (e.g. portfolio beta of 1.0). However, it doesn’t actually help us attain relative portfolio outperformance - since identifying factor correlation alone doesn’t tell you which way those factors will swing. All it helps with is portfolio management (i.e. improving beta) - it doesn’t really help you with outperformance attributable to fund manager skill (i.e. improving alpha).
This is where Value Investing comes in. As we’ve mentioned earlier, value investors like Warren Buffett and Charlie Munger tend to assign lower importance to the concept of Volatility As Risk (VAR) - since they do not consider portfolio volatility as Actual Investment Risk (defined as the chance of permanent loss of capital). As a result, their investment portfolios tend to exhibit much higher volatility than practitioners of more traditional portfolio management styles. If you recall from the title of this article, this is the actual problem that we’re trying to solve here.
In order to continue this explanation, we need to make an assumption for brevity’s sake. Firstly, let us assume that the value investor does indeed have higher alpha than his more contemporary peers - i.e. he is able to achieve better fundamental outperformance by picking stocks. To be clear, the only reason we are making this assumption is because there is no point in continuing on this train of thought if we assume that the value investor cannot even outperform on a fundamental basis. After all, the entire point of being a fund manager is to attain competitive outperformance - value investor or not - so I don’t think this is a particularly demanding assumption.
Now that we have assumed that the value investor is indeed able to achieve higher alpha than his contemporary peers (e.g. perhaps because he is only concerned with fundamental outperformance), the follow-up question to ask is then - how does this allow him to practice a Low-Volatility Value Investing Portfolio?
Firstly, he can put into practice the same low-volatility portfolio as aforementioned in the section above. By pairing different factors with varying correlations to each other, he should at least be capable of producing a porfolio composition with an ex-ante portfolio beta of 1.0 to the benchmark index. While this may not necessarily result in dampened ex-post portfolio volatility (due to unforseeable future circumstances), the value investor should at the very least be able to convince his fiduciaries that he has made a best effort to put a leash on portfolio volatility.
To provide an example, let’s use the aforementioned oversimplified scenario to illustrate. Suppose the value investor intends to neutralize interest rate risk, and attempted to do so by constructing a portfolio which is evenly split between the Financial and Real Estate sectors. Since in theory one sector should move upwards while the other should move downwards - regardless of which way interest rates actually swing - their respective impact on the aggregate portfolio with respect to interest rate volatility should cancel each other out - hence neutralizing interest rate risk.
Subsequently, the value investor (who we are assuming possesses superior alpha in this example) can select the stock within the Financial sector which he expects to outperform the most, and likewise pick the best-performing stock within the Real Estate sector. Since we are assuming that he actually has superior alpha, they should both outperform their respective sector peers regardless of the wider interest rate trajectory. And by virtue of being negatively correlated to each other at the sector level, both positions should also dampen portfolio volatility with respect to interest rate swings.
In this way, the value investor possesses a fund which in theory should possess dampened volatility (i.e. low beta) in the context of interest rate movements - while also possessing high alpha due to superior stock picks within the two sectors he has allowed himself to traverse within. If we were to extrapolate this practice across all factors (e.g. sectors, regions, etc) across the entire stock universe, the value investor could in theory construct a portfolio with an ex-ante portfolio beta of 1.0 (i.e. low volatility) by mimicking the factor exposure of the benchmark index. Then, by picking the best-performing stocks within the universe of stocks which that pre-determined factor exposure allows, he can subsequently deliver outperformance via superior alpha (i.e. stock picking skill).
There are two notable observations which can be made here. Firstly, since the value investor’s stock selection universe has now been reduced by that pre-determined factor exposure, he will likely be unable to match his original outperformance (i.e. if he wasn’t concerned with portfolio volatility). This is because while a volatility-agnostic value investor might decide to allocate more heavily towards a better return-inducing portfolio composition (e.g. more concentrated stock positions, more weighted towards certain sectors), he can now only pick stocks within the constraints of that aforementioned top-down portfolio strategy designed to dampen portfolio volatility (e.g. only within the Financial and Real Estate sectors). Hence, if a stock he really likes falls outside either the Financial or Real Estate sectors in the above example, he will be unable to participate in that stock position due to the artificial portfolio constraints he has shackled himself to.
However, the second observation that can be made is that despite these constraints, the value investor can still reliably outperform the benchmark index - as long as he possesses superior alpha. By attempting to construct a portfolio with an ex-ante portfolio beta of 1.0, he can in theory ignore the beta exposure of systematic risks on his portfolio and focus exclusively on achieving fundamental outperformance - i.e. a value investing strategy. This should in theory allow the value investor to not have to worry about macroeconomic developments - since his portfolio beta of 1.0 should have the same macro exposure as the benchmark index - and allocate his focus purely towards the unadulterated pursuit of fundamental outperformance. In this way, he can practice a Low-Volatility Value Investing Portfolio.
Conclusion
In doing so, the value investor - whose claim to outperformance rests solely on his ability to deliver superior alpha contingent on fundamentals-based outperformance - is not only able to deliver on the promise of his namesake, but at the same time also appease the majority of market participants whose preferences are significantly weighted towards participating in a low-volatility portfolio.
While it’s true that he may not achieve the same level of long-term outperformance as he would if he were practicing a truly volatility-agnostic approach, he is also able to remain faithful to a value investing philosophy (defined as fundamentals-based investing with superior risk:reward) with a crowd that may not necessarily prioritize long-term returns over short-term volatility.
There are several benefits to adopting such a Low-Volatility Value Investing portfolio strategy. On the portfolio front, this approach adds value to fiduciaries who may be intrigued in investing in a fund which practices a value investing philosophy - but are still unsure if they have the stomach for a volatility-agnostic approach. While the upside from such a Low-Volatility Value Investing strategy may be somewhat hamstrung as compared to a “pure” value investing approach (which is not constrained by volatility considerations), it should still be able to deliver superior Sharpe performance (i.e. return / volatility) - which may lend itself better to ancillary business considerations, such as for marketing purposes or expanding the acceptable investor base (e.g. attracting institutional investors). The sum of the above can even potentially end up yielding greater rewards for the fund manager himself - due to being able to attract a larger AUM base.
However, on the investment philosophy front, I think the real value-add here is that it makes value investing a palatable option for the vast majority of market participants who are highly concerned with portfolio volatility. As the recent drawdown in the Tech sector has demonstrated, many market participants who experience >20% drawdowns over a very short period of time simply do not have the appropriate temperament for a volatility-agnostic approach - which “pure” value investing tends to require. Also, as much empirical research has shown, over 50% of active fund management underperformance over the past decade can be attributed to investor redemptions occurring during the wrong times (i.e. selling low) - which is something largely outside of the fund manager’s control. This low-volatility value investing portfolio strategy allows these investors to stay invested when they otherwise wouldn’t - while also allowing them to participate in all the benefits of value investing, which its ardent fans have long since come to recognize.
While this low-volatility value investing portfolio strategy may not be too unfamiliar to most fund managers at the institutional level, it is my personal observation that many investors outside of the institutional space (including the sophisticated UHNW crowd) are not aware that such an option exists for them. As I have expounded on in the past in my super-long articles justifying the merits of value investing (linked below), value investing as properly practiced by Graham and Buffett can be likened to the PhD of equity investing - and making it accessible to more people can only be a good thing. This is what this article aims to do, and I hope it helps you take a step closer towards becoming a value investor as well.
Read my previous articles about Value Investing by clicking the links below!
If you like what you’ve just read, consider supporting me with an annual subscription at just $9.90/month for 2x research reports per month - so that I can deliver more attractive ASEAN stock ideas to your portfolio!
Click here to check out some of our other similarly attractive ASEAN equity research reports - only at Value Investing Substack, the premier place for ASEAN value investing stock ideas!
Type your email below and hit ‘Subscribe’ - to get deep-dives of ASEAN stocks & daily market news delivered to you… starting today!
