The myth of small-cap outperformance

“Small-cap stocks tend to outperform big-caps,” asserts Investopedia (“Understanding Small-Cap and Big-Cap Stocks,” 4 August 2022), “ ... because they are able to grow more rapidly than larger companies.” It’s an oft-made claim. “Small-caps do tend to carry more risk,” a funds manager conceded to CNBC, “but they should over time reward investors for taking that risk, meaning they normally outperform over long periods of time” (“Small-Cap Stocks Are Primed to Outperform Large-caps over the Next 10 Years,” 9 October 2019).

Are Investopedia and CNBC – and the many others who routinely advance such allegations – correct? By “small-caps” they typically mean American small-caps. Do their Australian counterparts outpace this country’s mid-caps and large-caps? The answer to both questions is clearly “no.” In this wire, I establish five sets of points:

1. Over short, medium and long terms since 2004, Australian small-caps haven’t outperformed mid-caps and large-caps. Quite the contrary: they’ve usually and cumulatively greatly (but not constantly) underperformed.
2. Not only are Australian small-caps’ average returns much lower: they’re also more volatile – and thus riskier by mainstream academic standards.
3. In the U.S., no small-cap premium has existed continuously or even generally since 1987. Over 12-month, five-year and ten-year intervals, American small-caps have matched the S&P 500 (over 20-year intervals, however, they’ve tended to outperform).
4. Why, despite strong evidence to the contrary, do many people mistakenly believe that Aussie small-caps excel? I propose two possible reasons.
5. Is now a good time to buy small-caps? Earlier this month, one funds manager asserted that it is; I demonstrate that it isn’t. More generally, it’s seldom a good time to buy Aussie small-caps.

A Key Preliminary 

Suppose that the price of the shares of an Australian company, X Ltd, is currently $1. You analyse its financial statements and, using conservative assumptions, conclude that they’re worth $2. You then observe that, according to my definition in the next section, X Ltd is a “small cap.”

I’m certainly not saying, nor even implying, that you shouldn’t buy these shares merely because X is a small-cap. If Leithner & Company’s estimated value of an individual company’s shares greatly exceeds their price, then we invest.

Since our formation in 1999, however, we’ve increasingly avoided small-cap stocks. That’s because they’ve mostly and usually offered mediocre and often poor value, and their overall returns prove it. Yet there are always a few diamonds in the rough, and we also seek to own individual underpriced securities regardless of their capitalisation. Value, in other words, trumps size.

What’s a “Small-Cap”?

What is a small-capitalisation stock? If it’s one of components of the S&P/ASX Small Ordinaries Index (“SOI”), then I regard it as a small-cap. “This index,” says Standard & Poor’s, “is designed to measure companies included in the S&P/ASX 300, but not in the S&P/ASX 100.” More generally, I designate the top-20 (ranked by market cap) of the ca. 2,000 companies listed on the ASX as “large-caps,” those ranked 21-100 as “mid-caps,” those ranked 101-300 as “small-caps” and the remainder (roughly 1,700) as micro caps. In the analysis below I’ll refer to the S&P/ASX 20 index as the “large-cap index” and the S&P/ASX 100 as the “mid-cap index.”

My conception of mid-caps is inexact because all of the components of the large-cap index are also members of the mid-cap index. The small-cap and mid-cap indexes, however, as well as the small-cap and large-cap indexes, have no members in common. On 30 June, the SOI’s total market cap was $292 billion; its largest stock’s cap exceeded $6 billion; its median component’s cap was $1.1 billion and its smallest member’s was $124 million.

Aussie Small-Caps’ Underperformance

Let’s say that in June 2004 you invested in three portfolios: a small-cap one (which tracked the SOI), a mid-cap one (which mirrored the ASX 100) and a large-cap one (which reflected the ASX 20). Let‘s also assume that since then you’ve rebalanced these portfolios quarterly (after the market’s close on the third Friday of March, June, September, and December), and that you’ve ignored dividends and paid no brokerage and management fees or taxes. Using monthly observations, Figure 1a plots the course per $1 of these three investments over the past 19 years.

Figure 1a: Growth per $1 (ex-Dividends, Fees and Taxes) in Three Indexes, Monthly Observations, June 2004-June 2023

Before the Global Financial Crisis, the small-cap portfolio did very well: during the three years to October 2007, not only did it more than double (to $2.13); it handily outperformed large-caps ($1.76) and mid-caps ($1.82). But then came the GFC: by February 2009, it collapsed 63% to just $0.79. In contrast, the large-caps and mid-caps lost less (34% and 39% respectively). Moreover, since then – and particularly since 2013 – the small-caps have greatly underperformed.

As a result, over the entire 19-year interval since June 2004, each $1 invested in the ASX 20 grew to $2.09 in June 2023. That’s lower than its peak before the GFC. Each $1 invested in the ASX 100 grew to $2.19, which exceeds its pre-GFC peak.

Each $1 invested in the SOI, however, grew to just $1.45. That’s not just lower than its pre-GFC peak; it’s also below the mid-caps’ and large-caps’ cumulative results. For almost 20 years, the small-caps have clearly – and cumulatively greatly – underperformed.

It gets even worse for owners of small-caps. The investment of $1 in June 2004 first grew to $1.45 in February 2006. Accordingly, over the past 17 years the investment’s capital has fluctuated – sometimes greatly – but cumulatively hasn’t grown a cent.

The SOI has underperformed over ALL of the intervals to June 2023 – ranging from one to 19 years – enumerated in Table 1. Not merely has no small-cap premium existed since 2004: if anything, there’s been a mid-cap premium.

Table 1: Compound Annual Growth Rates, Three Indexes, June 2004-June 2023

Note in particular the SOI’s loss during the most recent 15-year interval. Its capital return has been (1 -0.006)¹⁵ = 0.914. The market value of each $1 invested in June 2008, not including dividends – or brokerage, management fees and taxes – fell to $0.914 in June 2023. Talk about water torture!

Obviously, Australian small-caps haven’t underperformed during every interval since 2004. Importantly, however, on average they have.

I’ve computed all one-month (i.e., June 2004-June 2005, July 2004-July 2005, ... June 2022-June 2023) changes of each index, expressed them as compound annual growth rates and calculated these CAGRs’ mean; I’ve done the same for all 5-year and 10-year CAGRs; Table 1 summarises and Figure 1b plots the results.

Figure 1b: Average CAGRs, Three Indexes, June 2004-June 2023

On average, the SOI greatly underperforms in all intervals; moreover – and crucially – the longer is the timeframe the more worse is small-caps’ performance in both absolute and relative terms.

The small-caps’ long-term return – an average ten-year CAGR of just 0.2% – is particularly damning. During the average ten-year interval, the Small Ordinaries Index’s capital return has been a mere (1.002)¹⁰ = 1.02. On average after all ten-year intervals since June 2004, each $1 invested in the Small Ordinaries Index, not including dividends – or brokerage, management fees and taxes – has increased to a derisory $1.02. Aussie small-caps’ long-term returns make term deposits look like high-return investments!

The disparity between small-caps and the rest is likely even bigger than Figure 1a-b and Table 1 demonstrate. Mid-caps and large-caps’ dividends typically exceed small-caps’; accordingly, on a total return basis small-caps likely lag even further behind mid-caps and large-caps.

Not only do Australian small-caps comprehensively underperform: their returns are more variable than mid-caps’ and large-caps’. That is, according to the conventional conception of risk, which defines it as the standard deviation of a stock’s or index’s return, small-caps are riskier than mid-caps and large-caps.

As Figure 1c demonstrates, for small-caps, mid-caps and large-caps, short-term (one-year) CAGRs fluctuate much more than medium (five-year) and long term (ten-year) CAGRs. Yet small-caps are riskier than mid-caps and large-caps over all intervals, particularly over 12-month intervals.

Figure 1c: Standard Deviations of CAGRs, Three Indexes, June 2004-June 2023

What about the U.S.?

The Russell 2000 Index is American small-caps’ benchmark. It comprises the smallest (by market cap) 2,000 stocks in the Russell 3000 Index. The Russell 3000, in turn, is a float-adjusted cap-weighted index that seeks to benchmark all (and represents more than 97%) of America’s publicly-listed companies. The Russell 2000 and 3000 were devised and are maintained by FTSE Russell, a subsidiary of the London Stock Exchange Group. On 30 June 2023, the market cap of the median Russell 2000 stock was slightly more than $0.95 billion, and the Index’s total market cap exceeded $2.2 trillion (these and the other amounts in this section are $US).

The Standard and Poor’s 500 (“S&P 500”) Index is a float-adjusted and cap-weighted index which tracks the performance of approximately 500 largest companies listed on stock exchanges in the U.S. On 30 June 2023, the average market cap of the 503 companies currently comprising the Index was $77.8 billion, and the Index’s total market cap exceeded $39 trillion. S&P Global Inc., an American publicly-traded corporation that’s a component of the Index, devised and maintains the S&P 500.

Let’s say that in October 1987 you invested in portfolios that closely tracked the Russell 2000 and S&P 500 indexes. Let‘s also say that every year on the fourth Friday of June you rebalanced the portfolio that mimics the Russell 2000, and that every quarter (after the market’s close on the third Friday in March, June, September and December) you rebalanced the portfolio that mimics the S&P 500; you also continue to ignore dividends, brokerage and management fees, and taxes. Using monthly observations, Figure 2a plots the growth of capital over time per $1 invested.

Figure 2a: Growth per $1 (ex-Dividends, Fees and Taxes) in Two Indexes, Monthly Observations, October 1987-June 2023

(Commencing the series in the immediate wake of the Crash of 1987 doesn’t – with one exception – significantly affect the results below. It boosts the results (12-months, ... 35 years) that begin in October 1987 – and should thereby, according to small-cap enthusiasts, flatter small-caps’ relative returns.) Each $1 invested in October 1987 in the Russell 2000, grew to $15.42 in June 2023. That’s a CAGR of 8.0%. Each $1 invested in the S&P 500 grew to $17.19. That’s a CAGR of 8.3%.

As in Australia, so too in the U.S.: the Russell 2000 has underperformed over ALL of the intervals to the present enumerated in Table 2. As it is here, so it is there: no small-cap premium has existed since 1987; if anything, and again as in Australia, a mid-cap premium has existed.

Table 2: CAGRs, Two Indexes, October 1987-June 2023

Further significant points arise from a comparison of Tables 1 and 2. First, over every interval the Russell 2000 outperforms the Small Ordinaries. Second, over every interval the S&P 500 outperforms the ASX 20 and ASX 100. Finally, the American “small-cap” index outperforms both the mid-cap and large-cap Australian indexes.

Regardless of capitalisation and over short, medium and long terms, American stocks have outpaced Australian stocks. Adding dividends and computing total returns would greatly reduce but not eliminate the disparity.

Another crucial point: the Russell 2000’s underperformance vis-à-vis the S&P 500 is a recent development. Figure 2b plots the two indexes’ average CAGRs in all one-year, five-fear, 10-year and 20-year intervals since October 1987. On average the two indexes’ one-year and five-year CAGRs are identical (9.5% and 8.3% respectively); further, the Russell 2000’s 10-year and 20-year CAGRs slightly exceed the S&P 500’s (7.8% versus 7.6% and 7.2% versus 6.2% respectively).

Clearly, it’s not that small-caps per se underperform; there’s something about Australia’s that causes them to lag their American and Australian mid-cap and large-cap counterparts.

Identifying this discrepancy is relatively easy; explaining it is much harder. Certainly part of the answer is the innately higher valuation (which isn’t the same as value) of American vis-à-vis Australian equities. Another part, I suspect, is some American small-caps’ relatively large size – and most Australian small-caps’ relatively small size: the top one-third (by market cap) of the companies comprising the Russell 2000 on 30 June 2023 would qualify as mid-caps if they were listed on the ASX.

Figure 2b: Average CAGRs, Two Indexes, October 1987-June 2023

Why do Australian small-caps underperform? According to Cameron McCormack (“Why Aussie small caps are consistent underperformers,” Firstlinks, 14 June 2023), over the past 10 and 20 years, “global small-caps’ EPS growth (has) outpaced Australia’s by more than three times.” Moreover, the SOI’s percentage of non-profitable companies is twice the global average.

It’s not just Australian small-caps which underperform relative to mid-caps and large-caps: all Australian equities, regardless of capitalisation, lag relative to American equities. Comparison of Tables 1 and 2 shows that the Russell 2000 greatly outperforms the S&P/ASX 20 and 100 over one-year periods (CAGRs of 9.5%, 4.9% and 4.9% respectively), five-year periods (8.3% versus 2.6% and 2.5%) and 10-year periods (7.8% versus 3.0% and 2.8%).

Comparing CAGRs since June 2004 lessens but doesn’t eliminate the disparity: the Russell 2000 continues to outperform the S&P/ASX 20 and 100 over all one-year periods (CAGRs of 8.5%, 4.9% and 4.9% respectively), all five-year periods (7.2% versus 2.6% and 2.5%) and all 10-year periods (7.1% versus 3.0% and 2.8% respectively). Adding dividends into the calculations further lessens the disparity (during the past few decades Australian dividend yields have usually been significantly higher than American yields) but still doesn’t eliminate it.

What about the volatility of American small-caps’ returns? In this respect, too, they differ greatly from Australian small-caps: the variability of the Russell 2000’s one-year returns exceeds the S&P 500’s, but the reverse is true for five-year, ten-year and 20-year CAGRs (Figure 2c).

Figure 2c: Standard Deviations of CAGRs, Two Indexes, October 1987-June 2023

Why Do So Many Believe that Aussie Small-Caps Outperform?

Two sets of reasons come to mind. First, like many of the punters who frequent casinos, so-called investors (“speculators” is more apposite) don’t know that Aussie small-caps underperform relatively and absolutely. Perhaps that’s because people who should know, such as advisors and funds managers, don’t know or won’t say; or perhaps the punters or their advisers and funds managers do know but nonetheless reckon that they can beat the poor odds.

Figure 1b showed that the SOI’s average 12-month CAGR is 3.6% and that the ASX 20’s is 4.9%. The small-cap index’s average 12-month underperformance vis-à-vis the large-cap index is thus 3.6% - 4.9% = -1.3%. Figure 3 plots the SOI’s average underperformance relative to the mid-cap and large-cap indexes over all 12-month, five-year and ten-year intervals since June 2004.

“Investing” in Australian small-caps is like gambling in casinos: the longer you play, the more you’ll underperform.

Figure 3: SOI’s Average Net Performance over Three Intervals, June 2004-June 2023

A second key reason might explain why many people apparently believe that Aussie small-caps outperform – or, at least, that they can beat the poor odds. Like players in casinos, gamblers on Aussie small-caps occasionally, greatly but temporarily outperform. Conveniently overlooking or forgetting that on average they lose, they conclude that they can often win.

Figure 4 plots the SOI’s average performance relative to the ASX 100over all 12-month, five-year and ten-year intervals since June 2004. Positive numbers denote outperformance and negative numbers indicate underperformance.

Figure 4: SOI’s Performance Relative to the ASX 100, Three Intervals, June 2004-June 2023

The SOI’s average underperformance vis-à-vis the ASX 100 over all ten-year intervals, as Figure 3 showed, is -2.8%. Figure 4 indicates that its actual 10-year underperformance varies relatively little (its standard deviation is 1.5%). The SOI’s average underperformance over all five-year intervals is -1.9%, and its actual five-year underperformance varies somewhat more (its standard deviation, as Figure 1c indicated, is 3.7%).

For an extended period of time – the four years from April 2018 to April 2022 – the SOI’s five-year CAGR significantly (by an average of 2.27%) outperformed the ASX 100’s. Perhaps that anomalous period was long enough and the outperformance big enough to convince punters – many probably required no prompting, and never mind that since 2018 the five-year CAGR has become ever more negative – that small-caps were worth a flutter.

The SOI’s 12-month performance relative to the ASX 100’s fluctuates greatly: its mean, as Figure 3 indicates, is -1.3% but its standard deviation (as Figure 1c showed) is 19.9%. As a result, and as Figure 4 demonstrates, the SOI often outperforms – occasionally hugely but always temporarily. Before the GFC, it outperformed by as much as 40% per year (and by an average of 24%); it also greatly outperformed (by almost 30%) in the second half 2016, by 15-20% in late-2017 and early-2018, etc. Of course, it also underperformed disastrously – by as much as 40% – during the GFC and in 2013.

Small-cap speculators, I suspect, tend to remember the short periods of outperformance, discount or forget the long periods of underperformance, and are thus oblivious to the average – not just that the SOI underperforms the mid-cap and large-cap indexes, but also that over long stretches it can generate losses.

Is Now a Good Time to Buy Small-Caps?

One recent contributor to Livewire thinks so. In “The best time to invest in small caps is now” (3 July), Roger Montgomery writes: “Fears of recession, a market correction and runaway inflation have caused small-cap stocks to materially underperform their large-cap cousins. Those same fears have led many investors (to ask) how wise it is to invest in a small-cap fund now ... I believe you should be a contrarian investor and invest in a small-cap fund now.”

It’s unclear whether Montgomery’s referring to American or Australian small-caps; either way, he’s right about the underperformance. Yet he’s wrong about the reason: both countries’ small-caps have underperformed for decades – and thus regardless of ephemeral things such as fears of recession or the degree of consumer price inflation, etc. And his allegedly “contrarian” advice to invest in small-cap fund is demonstrably mistaken.

Table 3: SOI’s Performance Relative to S&P/ASX 100, Previous and Subsequent 12-Month Periods, June 2004-June 2023

I ranked the observations plotted in Figure 3 according to the SOI’s 12-month relative performance. I then divided these observations into five quintiles (i.e., ranked piles of observations, each of which, net of rounding, contains the same number of observations). Next, for each observation I computed the SOI’s CAGR relative to the S&P/ASX 100’s during the next 12 months. Finally, for each quintile I computed the SOI’s average absolute and average relative performance during the next 12 months. Table 3 summarises the results.

The SOI’s performance relative to the mid-cap index’s is strongly mean-regressive; in other words, the higher it rises above its mean at one point in time, the stronger is the likelihood that it subsequently falls (“regresses”) towards its mean, and vice versa. But its relative performance is only mildly mean-regressive for 12-month intervals since 2004. The more the SOI underperforms during a given 12-month period, as in Quintiles 1 and 2 of Table 3, the more it tends to outperform during the next 12 months. But the outperformance isn’t great – on average, it’s just (2.6% + 1.1%) ÷ 2 = 1.9%. Conversely, the greater is the SOI’s outperformance during a 12-month period, as in Quintiles 4 and 5, the more it tends to underperform during the next 12 months (by an average of 3.5%).

During the 12 months to 30 June, the SOI rose 5.2% and the ASX 100 9.9%; accordingly, the SOI’s underperformance was 5.2% - 9.9% = -4.7% (which is a huge improvement from the -19% in the 12 months to November of last year). That’s on the boundary between Quintiles 3 and 4. Accordingly, if past is prologue then during the 12 months to 30 June 2024 we can expect the SOI to underperform the mid-cap index by ca. (1.1% + 3.2%) ÷ 2 = 2.2%.

Of course, little in financial markets is certain; in particular, statements about the future are notoriously prone to error. So in fairness it’s important to acknowledge that I could be wrong and Montgomery could be right. Above all, given the huge variability of the SOI’s 12-month returns since 2004, during the next 12 months it could soar. But acting on the anticipation of a chance result is akin to the gambler betting that the next spin of the roulette wheel will recoup his recent losses: the odds are against it.

The SOI’s average underperformance over periods of one year, five years and 10 years tells us that it’s seldom a good time to invest in small caps as a whole. Table 3 tells us more. How to underperform systematically? Buy the SOI on those infrequent occasions when it’s greatly outperformed during the previous 12-months. How potentially and temporarily to outperform? Buy the SOI when its underperformance vis-à-vis the mid-cap and large-cap indexes reaches an extreme. The nadir of the GFC was such a time; today isn’t.

Conclusions and Implications

“There are lots of reasons why investors get excited about small (capitalisation) stocks, particularly in Australia,” a prominent asset manager stated in 2017. These reasons include “the higher growth rates that smaller companies can achieve” and “the myriad of successful individual stock stories that abound.” Many people dream about – and some repeatedly try – hitching their wagons to “the next Microsoft” (or Amazon, Apple, etc.) and riding it effortlessly to a fortune. For this or other reasons, “many investors in Australia are ... structurally overweight small-cap stocks ...”

If that’s still true, then many Australian advisors have been misadvising their clients – or else “investors” and advisors alike remain misinformed and overconfident.

In particular, they apparently believe that outside the large-caps and mid-caps, where analysts constantly assess vast amounts of information and thereby price these stocks accordingly, opportunities beckon. Relatively few and sometimes no analysts, they plausibly allege, research particular small companies; therefore, they invalidly infer, enticing prospects exist to produce superior returns.

The reality is very much otherwise: Australian small-caps usually, and cumulatively greatly, underperform mid-caps and large-caps; they lag over short, medium and long terms; they’re also more volatile (“riskier”). “Investing” in Aussie small-caps is like gambling in casinos: the more often and longer you play, and the more you wager, the more you’ll underperform the house.

My results corroborate research in the U.S. and globally. Vitali Kalesnik and Noah Beck (“Busting the Myth about Size,” Simply Stated, November 2014) provide an excellent – that is, concise, readable and reflecting valid and reliable research – overview.

Among their key insights: for any assertion that small-cap stocks outperform large-caps (e.g., they grow faster and have greater scope to grow), there’s one explaining why the reverse is true (e.g., they’re less financially sound and much more likely to fail and be delisted than mid-caps and large-caps).

Kaleshnik’s and Beck’s conclusion remains valid: more than 40 years after the initial publication of research purporting to uncover a small-cap premium, “the empirical evidence is extremely weak even before adjusting for biases.” In the U.S., the premium “is driven by extreme outliers (and the improper handling of data) which occurred three-quarters of a century ago ... Adjusting for biases, most notably the delisting bias, makes the (small-cap) premium vanish.”

So much for small-cap stocks and indexes; what about small-cap funds? 

1. Richard Ennis and Michael Sebastian of Ennis Knupp Associates, one of America’s largest pension consulting firms, assembled a sample of 128 small-cap managers. Adjusting it for management fees and survivorship bias, etc., outperformance virtually disappeared (“The Small-Cap Alpha Myth,” The Journal of Portfolio Management, Spring 2002, vol. 28, no. 3, pp. 11-16).
2. Based upon its analysis of the eVestment Database of small-cap managers, in 2016 Aon Hewitt, another large consulting firm, found that the median manager’s performance was worse for the 10-year period to 30 June 2015 than in its original analysis in 2001. Before adjustments, the median performance was less than 1% (originally around 4%). Adjusted for survivorship bias, liquidity and transaction costs, etc., the median result was negative.
3. “The Small-Cap ‘Alpha’ Myth” (Seeking Alpha, 10 May 2016) compared the average performance of all 479 actively managed small-cap funds (as classified by Morningstar) against benchmarks like the Russell 2000 and S&P Small-Cap 600. The average actively-managed small-cap fund underperformed the Russell 2000 Index by 0.24% per year net of fees.

Ken Fisher (The Little Book of Market Myths, John Wiley & Sons, 2013) is right: “outside just a few of the relatively hugest small-cap bounce periods, large-caps overall beat small-caps – and typically for agonizingly long periods. It can be mentally and emotionally trying (to invest in an asset class whose) relative payoffs are few and far between ... Those times when big stocks beat small are long enough to drive even the most patient (small-cap) investor absolutely insane.”

The bottom line: if you “invest” in small-caps, especially in Australia and whether directly or via a managed fund/ETF, over short intervals and like punters in a casino, you’ll occasionally and temporarily get lucky. Over longer periods, however, your results will fluctuate much less – and the odds are stacked strongly against you.

Click LIKE so that Livewire knows that you want more of this type of content. Click FOLLOW on my profile for notification when my next wire appears.