Quantitative and Systematic Investing: How Machines Are Changing Portfolio Management

The transformation of investment management from human judgment to mathematical models has been one of the defining trends of the past 30 years. At the pinnacle, Jim Simons' Renaissance Technologies has generated returns that no human stock-picker has come close to replicating, using mathematical models that have never been fully disclosed. At the accessible end, a private investor can now implement a simple momentum or value factor strategy through a UCITS ETF at 0.25% annual cost. Between these extremes lies a spectrum of quantitative approaches, each with its own logic, evidence base, and risk profile.

What Quantitative Investing Is

Quantitative investing uses mathematical models, statistical analysis, and data processing to make investment decisions — identifying what to buy and sell based on systematic rules rather than human discretion. The goal is to exploit patterns in financial data that are either (a) genuinely predictive of future returns, or (b) the result of market structure or behavioural biases that create persistent mispricings.

The rationale for removing human discretion is grounded in decades of behavioural finance research: human investors make systematic, predictable errors (overconfidence, loss aversion, anchoring, herding, availability bias) that consistent rule-following can avoid. If the rules are well-designed, a systematic process should outperform an inconsistent human applying the same principles imperfectly.

The key distinction from passive index investing is that quantitative strategies make active choices about which securities to hold, based on signals — they are active strategies implemented systematically rather than through human judgment.

The Quant Spectrum: From Simple to Complex

Rules-based factor strategies (smart beta) are the simplest form: systematic rules that deviate from market-cap weighting to capture documented return factors (value, momentum, quality, low volatility, size). These are covered in detail in our smart beta and factor investing guide. They are accessible through ETFs and are the most transparent form of quantitative investing.

Trend-following (CTA strategies). Commodity Trading Advisors — despite the name, they trade all asset classes — use systematic momentum signals to identify price trends in equity, bond, currency, and commodity markets. If equity markets are trending upward, a trend-following model takes a long position. If they turn downward, the model reverses. The key characteristic of trend-following is that it performs well precisely when markets make large, sustained moves — including bear market crashes. This makes it a genuine portfolio diversifier rather than a hidden equity correlation.

Statistical arbitrage. Stat arb strategies exploit statistical relationships between related securities — if two historically correlated stocks diverge, a stat arb model goes long the laggard and short the outperformer, betting on convergence. The strategy requires very fast execution, low transaction costs, and large-scale data processing. It is primarily the domain of institutional hedge funds.

Market-neutral quant strategies. By combining long positions (overweighted factors) and short positions (underweighted factors or short-sold securities), market-neutral quant funds aim for zero net equity exposure — returning purely the factor premiums without broad market beta. The AQR Absolute Return Fund and Man AHL TargetRisk Fund illustrate this approach.

Machine learning and AI-driven strategies. The frontier of quantitative investing applies deep learning, reinforcement learning, and natural language processing to financial data. These strategies seek to identify non-linear patterns that traditional statistical models miss.

The Quant Pioneers

Understanding the industry's history matters for assessing what systematic investing can realistically achieve:

Renaissance Technologies (founded by Jim Simons, a mathematician, in 1982) manages the Medallion Fund — the most successful investment fund in recorded financial history. Since 1988, the Medallion Fund has generated gross returns of approximately 66% per year — an extraordinary and unreplicated track record. The fund employs mathematicians, physicists, and computer scientists, not conventional finance professionals. Its methods have never been disclosed. The fund has been closed to outside investors since 1993 and is exclusively available to Renaissance employees. It is an existence proof of what quantitative investing can achieve at its best — but it is not a template that others have successfully replicated at this scale.

Two Sigma and D.E. Shaw are large, sophisticated quant hedge funds each managing tens of billions in assets (broadly $60–70bn apiece as of 2025), employing similar data-science-heavy approaches. They deliver strong but more conventional risk-adjusted returns than Medallion, with flagship strategies generally targeting double-digit annualised returns net of fees.

Man AHL and Winton Group are UK-headquartered systematic managers focused on trend-following and diversified systematic strategies. They are accessible to institutional and UHNW investors through managed accounts and UCITS funds. Man AHL has a track record dating to the 1980s.

AQR Capital Management (Cliff Asness) has been the most prominent advocate for factor investing and has published extensively on the academic foundations of the strategies it runs. AQR runs multi-factor strategies, risk parity approaches, and alternative risk premia. Its UCITS-wrapped funds are accessible to European investors.

Trend-Following: The Most Accessible Institutional Strategy

Of all systematic strategies, trend-following is the most accessible and most empirically well-supported for private investors. Its core portfolio benefit — positive performance during sustained market dislocations — makes it particularly valuable as a portfolio diversifier.

The evidence from 2008 and 2022 is instructive. In 2008, the SG Trend Index (a benchmark for trend-following CTA strategies) returned approximately +20% as equity markets fell 40-50%. In 2022, CTA strategies returned on average approximately +20-30% as equities fell 20-25% and bonds fell 10-20% simultaneously. These are exactly the scenarios in which conventional equity/bond portfolios provide insufficient diversification.

The trade-off is performance in choppy, trendless markets (2018, parts of 2011) where trend-following can produce meaningful drawdowns and periods of frustrating flat-to-negative returns.

Accessing trend-following in UCITS funds:

• Man AHL Trend Alternative UCITS Fund — the most prominent UK-accessible trend-following UCITS, with a long institutional track record
• Winton Fund UCITS — systematic diversified strategies
• AHL Evolution Fund — Man Group's machine learning-enhanced systematic strategy
• iM DBi Managed Futures Strategy ETF — US-listed, low-cost, replicates a trend-following return profile

Machine Learning and AI in Quantitative Investing (2024-2026)

The application of machine learning to financial markets has accelerated significantly with the availability of large-scale computing power, alternative data, and improved model architectures.

Alternative data — satellite imagery of industrial facilities, container ship traffic, retail car park occupancy, credit card transaction flows, social media sentiment analysis — is now used by major quant funds to generate signals before conventional reported data becomes available. MSCI and Bloomberg now sell data products that aggregate alternative data sources for institutional investors.

Natural language processing applied to earnings call transcripts, regulatory filings, and news flow identifies changes in management tone and sentiment that precede stock price moves. The technology is sophisticated and expensive; it is the domain of large quant funds rather than accessible retail products.

Reinforcement learning — where a model learns by trial and error in a simulated environment — is being applied to portfolio optimisation and trade execution. Practical applications remain largely in execution optimisation rather than alpha generation.

The critical caveat: despite enormous investment in machine learning applications to financial markets, there is limited public evidence that AI-driven strategies materially outperform well-designed conventional systematic strategies at the portfolio level over long periods. The most successful AI applications appear to be in high-frequency execution, alternative data processing, and risk management rather than in the core return generation that Renaissance Technologies achieved with simpler mathematical models.

The Risks of Systematic Strategies

Factor crowding. When a strategy becomes widely adopted, the expected premium narrows and the correlation between investors holding the same positions increases. The "quant quake" of August 2007 demonstrated this dramatically: multiple quant equity funds simultaneously experienced large losses as crowded positions in momentum and value factors unwound during a liquidity stress. The factor signals had not changed — but the crowded unwind created price impacts that the models had not anticipated.

Regime change. Systematic strategies are designed on historical data. If the underlying market structure changes — regulatory changes, market microstructure evolution, central bank intervention — the historical patterns that the model was trained on may no longer apply. The failure of many CTA strategies in the 2009-2012 period (as central bank intervention truncated natural market trends) illustrates this risk.

Overfitting. A quantitative model that is optimised on historical data may be capturing noise rather than genuine signal. If a researcher tests enough variations of a model on historical data, they will find combinations that appear to work remarkably well — but the apparent performance reflects data mining rather than genuine predictive relationships. Robust model validation and out-of-sample testing are important safeguards that are difficult to assess from outside the fund.

Correlation in crises. Many systematic strategies perform well in specific stress scenarios (trend-following in 2008) but may underperform in others (stat arb in the 2007 quant quake; factor strategies in COVID's initial crash). No strategy is a universal stress hedge.

Building Systematic Exposure in a Portfolio

For internationally mobile HNW investors, systematic strategies are most useful as genuine diversifiers — uncorrelated to equity market direction — rather than as equity substitutes.

A reasonable allocation framework:

• Core systematic exposure: 5-10% of portfolio in a diversified trend-following fund (Man AHL Trend UCITS or equivalent) as an equity diversifier
• Factor ETFs: as part of the equity allocation, replacing some market-cap-weighted equity exposure with multi-factor ETFs (iShares MSCI World Multifactor, AQR Multi-Strategy UCITS)
• Alternative risk premia funds: for UHNW investors with access to institutional share classes, diversified alternative risk premia funds (value, carry, momentum, defensive across asset classes) add genuine diversification

The primary test is whether the systematic allocation is genuinely uncorrelated with the rest of the portfolio in stress scenarios. Past crisis performance (2008, 2022) provides some evidence, but as always, past performance is not a guarantee of future behaviour.

How Global Investments Can Help

Systematic and quantitative strategies require more sophisticated due diligence than conventional equity or bond funds — understanding the underlying strategy logic, assessing the quality of risk controls, and evaluating whether historical performance reflects genuine skill rather than data mining. Our team has experience evaluating systematic fund managers and helps clients select appropriate exposure based on their portfolio needs. We help internationally mobile investors access institutional-quality systematic strategies through appropriately structured UCITS vehicles and assess how they fit within the broader portfolio context. Systematic strategies can experience significant drawdowns; past performance is not a guide to future results; investments can fall as well as rise; seek professional financial advice before investing in quantitative strategies.