The U.S. fairness market has periodically been dominated by a couple of technology-driven shares, most just lately the so-called “Magnificent Seven.” Traditionally, comparable dominance occurred through the Nifty Fifty period within the Nineteen Sixties–Seventies and the dot-com growth within the Nineties. These intervals of concentrated management usually led to momentary outperformance, however systematically capturing such features has confirmed difficult. Our research investigates the potential to take advantage of expertise sector dominance utilizing momentum-based methods throughout Fama–French 12 trade portfolios, analyzing whether or not long-only, long-short, and rolling-basis approaches can generate persistent alpha, and assessing the constraints of straightforward timing strategies.
We’re as soon as once more in a interval when the U.S. fairness market is dominated by a small group of shares — the so-called Magnificent Seven. This group consists of Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta Platforms, and Tesla, seven technology-oriented giants that collectively account for greater than 1 / 4 of the S&P 500’s complete market capitalization. Their affect on general market efficiency has turn into extraordinary, pushed by innovation in synthetic intelligence, cloud computing, and digital platforms.
This degree of focus just isn’t unprecedented. Within the late Nineties, expertise and telecom shares led the market to excessive valuations earlier than the dot-com correction. Within the Nineteen Sixties and Seventies, the Nifty Fifty represented an identical period of dominance by a handful of large-cap progress shares. Even earlier, vitality and industrial corporations performed a comparable function throughout postwar expansions.
For the reason that present management is once more pushed primarily by the expertise sector, our most important curiosity lies in understanding how dominant this sector actually is in contrast with the remainder of the market and whether or not such dominance might be systematically exploited.
For this evaluation, we utilized the dataset of 12 trade portfolios: Nondurable Items (NoDur), Sturdy Items (Durbl), Manufacturing (Manuf), Power (Enrgy), Chemical substances (Chems), Enterprise Gear (BusEq), Telecommunications (Telcm), Utilities (Utils), Retailers (Retailers), Well being (Hlth), Finance (Cash), and Different (Different), sourced from Fama-French. The dataset incorporates month-to-month observations for every portfolio from July 1926 to June 2025. Our proxy for the IT sector for the entire 100year interval is the Enterprise Gear (BusEq) sector/trade.
To seize the potential of the technological sector’s rising pattern, we concentrate on investing on this sphere individually, utilizing a number of completely different approaches, earlier than continuing to an examination of all 12 industries.
As a primary step, we outline the benchmarks, in opposition to which our buying and selling methods will likely be in contrast:
the common of all 12 trade portfolios,
the common of 11 trade portfolios excluding the expertise sector – Enterprise Gear (BusEq),
the expertise sector (Enterprise Gear – BusEq) alone.
Fundamental efficiency traits in tables are introduced as follows: the notation perf represents the annual return of the technique, st dev stands for the annual normal deviation, max dd is the utmost drawdown, adjusted Sharpe r is calculated because the ratio of perf to st dev and adjusted Calmar r because the ratio of perf to max dd.
What we are able to be taught concerning the efficiency of the technological sector if we research its efficiency in opposition to friends? Traditionally, we are able to establish 3 main intervals of the technological revolutions. The primary occurred within the Seventies, through the period of the Nifty Fifty and the rise of early computing giants akin to IBM, pushed by improvements like the primary microprocessors. Enterprise Gear sector primarily included producers of workplace and industrial equipment reasonably than trendy IT companies. Regardless of improvements just like the microprocessor, progress was restricted by stagflation and excessive rates of interest, leaving the sector’s efficiency in step with the broader market.
The second wave got here within the Nineties, marked by the web growth and the explosive progress of software program and telecom corporations main as much as the dot-com bubble. In the course of the Nineties, fairness markets have grown quickly as inflation fell, financial coverage stabilized, and monetary deregulation took maintain. The growth of index funds, derivatives, and computing expertise improved market effectivity and fueled an extended structural bull market by means of the Nineties, notably within the expertise sector. Nonetheless, this era was adopted by a noticeable decline in returns thereafter.
After the early-2000s downturn, the expertise sector stayed subdued till round 2010–2012, when cloud computing, smartphones, and platform-based fashions reignited progress. Firms akin to Apple, Amazon, Google, Microsoft, and Fb drove a brand new wave of growth by means of the 2010s. Lately, synthetic intelligence has added additional momentum, reinforcing the dominance of huge expertise companies in international fairness markets.
Nonetheless, historical past exhibits that after every of those revolutionary intervals, markets ultimately returned to actuality. The expertise sector, following phases of exuberant progress, tended to expertise substantial underperformance as valuations normalized and management rotated to different industries. Will the present bull market in AI shares finish the identical?
Our exploration of trade portfolios encompass a number of funding approaches with the goal of beating the benchmarks. We’ll current them one after the other.
Transferring common technique
As a primary step, we determined to design buying and selling methods primarily based on a transferring common. Firstly, we centered on the a part of the market fashioned by the expertise sector, or extra exactly, on how a lot of its efficiency might be defined by the remaining trade portfolios.
For every month, we subtracted the common return of the opposite eleven portfolios from the return of the expertise sector. In different phrases, we labored with the technology-vs.-rest unfold.
We then constructed an fairness curve from this unfold and calculated its transferring common. If the present worth of the unfold exceeded the common worth of the fairness curve over the required interval, the technique took an extended place in expertise shares and a brief place within the 11 different sectors/industries for the next month. In any other case, the positions have been reversed, lengthy within the 11 different sectors/industries and brief in expertise shares. When the present worth of the unfold is above its transferring common, it signifies that the expertise sector’s latest efficiency exceeds its long-term pattern, suggesting strengthening relative demand. In such conditions, traders could also be reallocating towards expertise, which may sign the continuation of its outperformance within the close to time period. We tried this strategy on completely different home windows for the transferring common, particularly from 3 months to 12 months. This process was repeated month-to-month. For higher visualisation is graphical interpretation in logarithmic scale.
From each the graph in Determine 4 and Desk 2, we are able to see that this strategy is reasonably worthwhile however extraordinarily risky. The intervals with out a clear pattern are lengthy, and the technique tends to earn a living solely when a expertise bubble is inflating or bursting. In any other case, its efficiency is reasonably weak and falls nicely in need of our expectations, which is why we moved on to a distinct methods.
Lengthy-only momentum technique
Subsequent, we centered on a momentum-based strategy. This time, we in contrast the momentum of the expertise portfolio with the momentum of the common of the remaining 11 trade portfolios. If the momentum of expertise was increased, we took an extended place in expertise for the next month. In any other case, we went lengthy within the common of the opposite industries. This process was repeated on month-to-month bases. As soon as once more, we examined varied momentum lengths, starting from 3 to 12 months . For higher visualisation, we use a logarithmic scale.
We will see that though the long-only methods obtain increased efficiency, they largely fail to outperform the benchmarks. The return-to-risk values are very comparable throughout completely different settings, and aside from the 12-month momentum, the place we barely outperform the benchmark, different MA lengths are unable to exceed benchmarks in each return-to-risk ratios concurrently. Subsequently, we determined to experiment additional with momentum and developed an extended–brief technique.
Lengthy-short momentum technique
On this case, we once more in contrast every month the momentum of the expertise portfolio with the momentum of the common of the remaining 11 portfolios. Nonetheless, this time, if the momentum of expertise was increased, we took an extended place in expertise and a brief place within the common of the 11 portfolios. In any other case, we reversed the positions, lengthy within the common and brief in expertise. Graphical outcomes are introduced in logarithmic scale.
In comparison with the transferring common methods proven in Determine 4 and Desk 2, the outcomes have improved barely, however relative to the earlier strategy and particularly to the long-only benchmarks, the efficiency is nothing spectacular. Subsequently, we determined to take a distinct perspective, because it turned out that fixing the best-performing trade upfront just isn’t an optimum strategy. Expertise just isn’t all the time the strongest sector, and the intervals when it performs nicely are comparatively few. Furthermore, this technique additionally introduces a look-ahead bias. Subsequently, it’s higher to design an algorithmic strategy that selects the best-performing trade relative to the others every interval, reasonably than assuming it’ll all the time be the expertise sector.
Lengthy-only rolling-basis technique
This rolling-basis technique can be momentum-based, however this time we evaluate the momentum of all 12 sectors/industries in opposition to one another. For the next month, we make investments solely within the sector with the very best momentum, successfully letting the info decide which sector is the most effective performer. As earlier than, we set the momentum calculation window to vary from 3 to 12 months and the method is repeated every month. For higher visualization, we once more use a logarithmic scale.
The outcomes of this technique recommend that this strategy is extra significant and that it certainly is sensible to “put money into the bubble.” Nonetheless, the efficiency and ratios figures are solely barely higher than the benchmarks. Subsequently, we moved on to a distinct strategy.
Lengthy-short Rolling-Foundation technique
On this case, we additionally examined an extended–brief strategy. Every month, we once more in contrast the momentum of all portfolios, however this time we went lengthy within the best-performing trade and brief within the remaining ones with equal weights for the next month. We apply this strategy primarily to look at the alpha, that’s, the portion of efficiency attributable to technique, reasonably than market actions. Once more, we repeated this course of every month, however this time we prolonged the momentum window, analyzing intervals starting from 1 month to 12 months. Once more, graphical interpretation is in logarithmic scale.
With this technique, we are able to already observe a transparent alpha, exhibiting that longer horizons (8–12 months) are more practical than shorter ones in figuring out the outperforming sector. When it comes to efficiency and ratio metrics, the outcomes don’t strategy the benchmarks; nonetheless, we didn’t anticipate that, as all methods are primarily based on long-short and subsequently all the metrics are often decrease than long-only variants.
We got here up with one final thought — to look at when this outperformance truly happens, that’s, when the best-performing trade begins to outpace the others. To discover this, we carried out a easy evaluation by creating common fairness curves of all lengthy–brief momentum methods after which constructing a technique, that’s timing this combination efficiency fairness curve.
Modified Lengthy-Quick Rolling-Foundation methods
Within the closing half, we determined to additional refine the lengthy–brief rolling-basis strategy and developed 3 new methods. For every month, we recognized the portfolio with the very best momentum throughout all time home windows, from 1-month to 12-month momentum. Within the following month, for every momentum horizon, we went lengthy within the best-performing sector and brief within the remaining ones, simply as within the earlier part. This time, nonetheless, we mixed all these approaches concurrently, making use of them equally-weighted and aggregating them right into a single composite technique. Let’s label this technique as “common mother“.
The second and third methods are derived from the primary one. After creating the fairness curve of the “common mother” technique, we first concentrate on its 12-month most. If, in a given month, the “common mother” fairness curve reaches its 12-month excessive, we put money into the “common mother” technique for the next month. In any other case, we keep out of the market. As typical, this course of is repeated every month. On this manner, we created the “12M Mother Max” technique.
The final examined technique may be very comparable, however as an alternative of specializing in the utmost, we take a look at the 12-month transferring common. If, within the present month, the worth of the “common mother” fairness curve is above its 12-month transferring common, we put money into the “common mother” technique for the next month. In any other case, we keep out of the market, repeating this course of every month. We label this technique “12M Mother MA“.
All 3 approaches are then plotted collectively alongside the benchmarks for comparability in logarithmic scale.
Reviewing the “common mother” technique, we are able to observe that there was alpha as much as the Fifties, however that is primarily based on Fama–French knowledge, and particular person sectors at the moment couldn’t be simply traded, so this represents a theoretical alpha. From the Fifties onward, many of the alpha efficiency happens in three distinct intervals: the Seventies, Nineties, and 2020s, akin to new technological eras. Exterior of those intervals, the efficiency of the “common mother” is roughly flat. Sadly, attempting to enhance the “common mother” through the use of the brand new excessive or transferring common filter strategy doesn’t enhance the efficiency of return-to-risk ratios. So, on the finish, we are able to conclude that the straightforward “purchase the most effective performing sector during the last 12-month interval, brief the remainder, maintain for 1 month” technique is the most effective strategy to take advantage of the cross-sectional sector efficiency dispersion.
Conclusion
Our evaluation exhibits that technology-driven intervals have dominated fairness markets, however systematically exploiting them stays difficult. Easy moving-average and momentum methods centered solely on expertise can sometimes generate outperformance, but they’re extremely risky and infrequently fail to surpass benchmarks. Rolling-basis approaches that choose the best-performing trade every month present clearer alpha, notably over longer momentum horizons, however efficiency stays modest exterior the main technological revolutions of the Seventies, Nineties, and 2020s.
Makes an attempt to time these intervals utilizing easy heuristics, akin to new highs or transferring averages, proved tough, underscoring the episodic and unpredictable nature of sector-driven outperformance. Whereas momentum methods can seize developments throughout technological waves, they need to be carried out with warning as they have a tendency to ship alpha solely over the longer horizons.
Writer: Sona Beluska, Junior Quant Analyst, Quantpedia
Are you in search of extra methods to examine? Join our publication or go to our Weblog or Screener.
Do you wish to be taught extra about Quantpedia Premium service? Verify how Quantpedia works, our mission and Premium pricing provide.
Do you wish to be taught extra about Quantpedia Professional service? Verify its description, watch movies, overview reporting capabilities and go to our pricing provide.
Are you in search of historic knowledge or backtesting platforms? Verify our checklist of Algo Buying and selling Reductions.
Would you want free entry to our companies? Then, open an account with Lightspeed and luxuriate in one 12 months of Quantpedia Premium without charge.
Or comply with us on:
Fb Group, Fb Web page, Twitter, Linkedin, Medium or Youtube
Share onLinkedInTwitterFacebookCheck with a buddy
