Far from normal

By anyone’s standards, 2025 has been an eventful year in equity markets. Probably dating back to the Trump election in the US in November 2024, and the uncertainty and volatility immediately following the inauguration, multiple regimes seem to have come and gone. 

Each seems to have been quite separate:

• The announcement of Deepseek in January 2025 shook certainty from the AI juggernaut when it was realised that larger and more intensive AI models may not be the answer.
• Dramatic risk aversion in the face of proposed tariffs and other disruptions to global trade from the US, culminating in “Liberation Day” in early April.
• Dramatic risk seeking when many of the Liberation tariffs were discovered to be simply negotiating tactics, or others could not be delivered at all, and were mostly put on hold. By June tariff arrangements with only 8 countries had been negotiated.
• A joint AI capex and equity market “bubble” which seems to stem from excessive extrapolation of the productivity gains from AI. This remains a serious problem(1).

The uncertainty prior to “Liberation Day” saw equities market sell off sharply, with a rotation to quality and away from uncertainty and volatility-exposed names. This reaction to US federal government policy gyrations was in anticipation of sharp increases in US inflation and a slowdown in global growth. Stocks with high levels of shorting or high beta were especially punished, while low beta and low levels of shorting were not. Any stocks or portfolios which were long quality and/or short uncertainty or volatility benefited unexpectedly. This alpha is best considered a windfall, because while fortuitous it would not have been part of the modelling of expected performance.

Since “Liberation Day”, and the lack of traction of these policies, the market has sharply reversed its risk preference and has been on an extended “junk” or risk rally. A few related factors seem to have contributed to this:

• Reversion of the (perhaps excessive) short selling in the run-up period.
• Resolution of the pre-Liberation Day policy uncertainty, as tariffs were paused or retracted into a more clearly defined “risky” future.
• The further concentration of the market into tech firms, and the dramatic upswing in projected data centre capex and energy/resource needs to support it.
• Strong retail flow, into both single stock equities and into ETFs. Retail flow tends to be both short term focussed and risk seeking.
• Index tracking flows have continued to grow sharply. These simply buy stocks at market cap weights, so larger stocks (largely tech) get a disproportionate proportion of the inflow, driving them higher.

Quality as an alpha factor has had a difficult time, as have any factors with a low beta or low volatility tilt. Most quant models are built knowing that a long-term tilt towards lower risk – usually through quality – pays off. For the second half of this year, however, this has not been the case, although high beta quality has been rewarded.

At RQI, we recognise that there is a persuasive long term alpha argument to tilt exposure to lower risk and higher quality stocks, but there are times when the rewards to this reverse. This has been one of those times. With no obvious catalyst for a return to “normal”, prudence suggests downweighting exposure to this in the interim.

Below we explore the nature of this rally, noting that it is:

• Highly skewed towards certain sectors, especially tech and the Mag 7.
• Most powerful among the most shorted stocks and the highest beta stocks, and the most shorted stocks are spread across many sectors (high beta stocks are not).
• Not a “junk” rally, in the sense that higher quality stocks (measured as ROE) have outperformed low ROE stocks. However, this is concentrated in expensive high beta firms in certain sectors.
• Not really a large cap rally – large and small stocks as separate groups have had similar returns. Large caps have outperformed a little.

Risk rally?

The uncertainty following the Trump election and inauguration, followed by the “Liberation Day” tariff announcements and unwinding, triggered wild swings in risk appetite which were reflected strongly in equity markets. Chart 1 below shows the cumulative total return of the MSCI World index from January 1, 2025, to December 1, 2025, along with carve out of other developed market countries (US, GB, CA, JP and AU).(2)

Chart 1: Cumulative Index Returns: MSCI World and MSCI World (US stocks only) (in AUD)

The sell off until Liberation Day is quite marked, as is the rebound. Note that in USD terms the market bottomed on April 8, 2025, but gyrations in the AUD/USD exchange rate meant that the AUD bottom was not reached until April 21, 2025. The rebound was very sharp, up from this point by approximately 25% by December 1, 2025. The US equity market sold off by more but has rebounded at about the same rate as the broader index.

The fear of catastrophic implications of US government policy drove the selloff, and the backdown from those policies seems to have driven the rebound. So, sharp risk-off, followed by sharp risk-on.

Table 1: Sector returns in US MSCI World stocks – before and after “Liberation Day”. 

Again, note returns are in AUD, so bottom of the market was actually April 21

Table 2: Mag 7 fall and bounce, before and after “Liberation Day”(3), plus % short at December 1 2025*(4).

Look out for the Shorts(5)

Chart 2 below shows the cumulative return to the most shorted US stocks in MSCI World. That is, we take the reported aggregate short position from the central reporting agency in the US (FINRA: Financial Industry Regulatory Authority) on behalf of the SEC(6). This data is reported every two weeks following collection from all customer and proprietary accounts. The data is necessarily lagged by approximately 2 weeks (between trade settlement date and publication date) so results will always be a little inaccurate. However, many stocks remain heavily shorted for extended periods, so the impact of the delay is lessened.

We then calculate the cumulative returns to a portfolio holding just these most shorted stocks (equally weighted) until the next short sales report, when the portfolio is reset. We define shorting as number of shares held short divided by the total shares on issue.

Chart 2: Cumulative Returns to Most Shorted US Stocks in MSCI World (in AUD)

Clearly, the more heavily shorted the stocks, the more they have bounced. However, more heavily shorted stocks were not sold off more than all other stocks prior to April 21.

What are these short stocks?

The most shorted 50 stocks are not the large tech (Mag 7) stocks of course, as we can see in Table 2 above. Their registers are too large. There are certainly many smaller cap stocks, and Information Technology dominates, however all sectors have more than one name. The top 50 stocks each fortnight yields 96 unique stocks from January 1 to December 1, spread across sectors as follows:

Table 3: Sector membership of top 50 shorts (US stocks, MSCI World)

Some stocks appear in every fortnightly report. For example, the following firms have consistently more than 10% of their register short throughout 2025. Performance has been mixed between January 1 and December 1:

• RocketLab (NASDAQ: RKLB) (not Rocket Companies), an aerospace company developing guidance systems (up 40%).
• Albemarle Corp, specialty chemical company, specialising in Lithium compounds (up 50%).
• International Paper Company (NASDAQ: IP), which creates renewable packaging and paper products (down 20%).
• Rivian (NASDAQ: RIVN), the electric vehicle firm (flat).
• IonQ (NASDAQ: IONQ), a quantum computing developer (flat).
• Reddit, news and entertainment website (up 20%).
• Bloom Energy (NASDAQ: BE), fuel cell technology company (up 290%).
• Moderna (NASDAQ: MRNA), RNA pharma company (down 40%).
• Enphase Energy (NASDAQ: ENPH), a global home energy technology company (down 60%).

Note in passing that on occasion the list contains stocks like Rocket Companies (NASDAQ: RKT), a mortgage provider, which had up to 45% of its shares shorted until its takeover of Mr Cooper in October 2025. This was probably part of a merger arbitrage trade (i.e., short bidder, long target). It has since reverted to only 4% of its shares shorted. It is up almost 100% between January 1 and December 1, 2025.

Beta, size, ROE and value

To underline the “abnormal” nature of the market moves we saw in 2025, we go further to look at four other characteristics: beta, size, quality (as ROE) and value (as EY (NTM)(7)). We look mostly at the US stocks in MSCI World.

Note that to avoid the impact of the Mag 7 and market cap in these charts, we have used equally weighted returns in each group rather than market cap weighted.

As the charts below, we find the following:

• High beta stocks dramatically underperformed low beta stocks prior to “Liberation Day” and then have dramatically outperformed ever since.
• The smallest US stocks underperformed prior to Liberation Day and have not outperformed large caps following it. We did not see this in the broader MSCI World benchmark, where large cap stocks only outperformed small caps slightly.
• What we have seen is very definitely a high ROE rally. Low ROE stocks have been flat since May 2025 following a small recovery. High ROE stocks (of which tech are a large part) have been very strong.
• Expensive stocks (measured as EY (NTM)) have also sharply outperformed, while cheap stocks have been flat.

Taken together, our summary is that expensive, high quality, high beta stocks have been the strongest performers. 

There is not a large difference between large and small cap stocks on average: large caps have grown sharply and slightly faster than small caps. Note that we see large clear sector differences in beta. For example, Consumer Staples has a very low beta, and its performance was mostly flat. Information Technology has a very high beta and its performance tracked down then sharply up.

It is not as clear cut in size, ROE and EY (NTM). The largest average sectors are Information Technology, Consumer Discretionary and Communications Services and these have performed strongest. The highest ROE sectors are Information Technology, Consumer Discretionary and Consumer Staples: tech and consumer discretionary have performed strongly. Similarly, the most expensive sectors are Information Technology, Industrials and REITs: only tech stands out.

Chart 3 shows the split between high and low beta stocks. As noted in Table 4, there are strong sector effects at play which colour these results. Tech stocks dominate the expensive high beta part, while utilities and consumer staples dominate the cheap low beta part. Taken with the sector performance in Table 1, we can see that the difference between high and low beta performance is also much the same as the difference between the performance of sectors.

Of course, in an upwardly moving market, we would always expect high beta stocks to outperform low beta stocks. It is the scale of this difference (more than 30% when the market is up about 15%) that is so remarkable. Again, the high beta nature of the Information Technology sector is important.

Table 4: Cap weighted average betas for sectors, US stocks in MSCI World

Chart 3: Cumulative Returns to 100 highest and 100 lowest beta US Stocks in MSCI World (in AUD), equally weighted.

Chart 4 separates the largest US stocks from the smallest (in MSCI World) and shows that the smallest stocks sold off much more in the run up to Liberation Day than the largest stocks, and both have rebounded similarly since then (the larger stocks have fared slightly better than the small stocks). While the smaller US stocks we speak of here are only the smallest in the MSCI World universe, and not a genuine small cap universe, note that the performance of the Russell 1000 (largest 1000 US stocks) outperformed the Russell 2000 (next largest 2000 stocks) by about 4% over the period from January 1 to December 1 (14.8% compared to 11%). This is not insignificant but in no way suggests it is the main driver of the abnormal returns we have seen.(8)

Chart 4: Cumulative Returns to 100 largest and 100 smallest US Stocks in MSCI World (in AUD), equally weighted

Chart 5 looks at quality, proxied by ROE. We plot the cumulative returns to the largest and smallest US stocks in MSCI World and see that while the sell-off is similar, the higher ROE (better quality) stocks rebound much more strongly(9). This shows that the rally we have seen is not a junk rally, as has been suggested elsewhere.

Chart 5: Cumulative Returns to 100 highest ROE and 100 lowest ROE US stocks in MSCI World (in AUD)

Finally, Chart 6 splits the US stocks with the MSCI World universe between expensive and good value, as measured by EY (NTM). Very clearly, expensive stocks have dramatically outperformed good value stocks, at least until November when they suffered a sizeable reversal.

Chart 6: Cumulative Returns to 100 most expensive and 100 best value US stocks in MSCI World (in AUD)

Meme stocks?

One last comment is on meme stocks. One explanation floated for this abnormal return behaviour has been retail investors chasing so-called “meme” stocks (stocks which attract the attention of risk seeking retail investors for speculative reasons but have little fundamental appeal).

It is difficult to define a meme stock in advance, but we can say that these stocks are speculative, usually tech oriented, and have a lottery-like payoff – in the sense that their chance of success is low, but they pay off handsomely when they are successful.

It is clear that this is not simply a meme stock rally, as it is broad based, with expensive high quality high beta stocks the main drivers. This does not mean that some stocks which have these characteristics have not performed well – for example, Opendoor Technologies (NASDAQ: OPEN) (an online residential property advertising site) ran from USD0.90 in July to over USD10 in September, before retreating to USD7.14 at December 1. Or GoPro Inc (NASDAQ: GPRO), which ran from USD0.80 in July to a peak of USD2.90 in September before falling back to USD1.65 by December 1. Even IonQ (NASDAQ: IONQ), mentioned above in the heavily shorted list, spiked from USD44 in September to over USD80 and then back by mid-November.

There are too few of these to drive an entire market in the way we have seen, and many stocks which received press as meme stocks (for example, Beyond Meat Inc NASDAQ: BYND)) have performed very badly.

What does all of this tell us about investment strategies?

Investment strategies which have style, sector or beta tilts during this period could have suffered wild swings in performance. A few examples:

• A strategy that was Value oriented might have suffered during this period due to the tailwind to expensive stocks
• This may have been ameliorated somewhat with a tilt towards better quality names with the Value universe as higher quality (ROE) outperformed.
• However, high quality stocks are often lower beta stocks – that is, less sensitive to market moves – so high quality and high beta is preferred. This again suggests a tilt to Information Technology.
• An aggregate bet against beta (for example, less than 1 in a benchmark aware strategy or negative in a market neutral) might have seen performance drag.
• These last two points are potentially also true in quant alpha models as they try to capture the many dimensions of quality. Low beta quality models (like low volatility, betting against beta, betting against uncertainty, …) would have suffered, but a tilt to highly profitable high beta stocks (tech again!) would have done well.
• A strategy that remained underweight tech, and especially underweight the Mag 7, would have found positive alpha difficult to come by. This story – market concentration and the difficulty of adding alpha – has been a theme for several years.
• Growth and tech-oriented investment management should have done well.

Conclusions

From the Trump electoral success in November 2024 until the announcement of so-called “reciprocal tariffs” on “Liberation Day” (early to mid-April 2025), the market was in turmoil, culminating in a dramatic sell-off and then policy capitulation from the Trump administration. This capitulation was taken well by the market, as a dramatic relief rally, where the market bounced hard and continued to run all the way through to the end of 2025. This coincided with (and perhaps was partly driven by) the ongoing AI and tech stock euphoria. Strong retail flow (risk-seeking) and continued growth of index funds have probably exacerbated this.

We have endeavoured, with some small success, to unpick this abnormal period and make some comments on likely drivers. What we have found suggests that it is:

• Concentrated in some sectors, especially tech, and in the Mag 7.
• Most powerful among the most shorted stocks and the highest beta stocks. The most shorted stocks are spread across many sectors, while the highest beta stocks are not.
• Not a “junk” rally, in the sense that higher quality stocks (measured as ROE) have outperformed low ROE stocks. However, this is concentrated in expensive high beta firms in certain sectors.
• Not specifically a large cap rally – large and small stocks as separate groups have had similar returns, although large cap has done a little better.

From our perspective, and that of our clients, it’s important to try to draw some conclusions on how different investment management techniques might have fared during this period. Only a tilt to high quality, expensive, high beta stocks (for example tech) would have had a tailwind. Most other more conservative styles, relying more on historical factor performance, might have struggled.

We have titled this paper “Far from Normal” to make the point that we live in interesting times, and market and factor performance is anomalous when compared to history. We would also contend that it’s hard to predict such events a priori(10), and that more normal factor performance – which have seen through many years of market study and have built investment strategies upon – should revert at some point. Trying to forecast when this happens is hard, but as we know “this time is different” are the four most dangerous words in the market.

Managed Fund

RQI Australian Value Fund - Class A

Australian Shares

View

Managed Fund

RQI Australian Small Cap Value – Class A

Australian Shares

View