After fifteen-plus years in equity research and conversations with hundreds of global investors, we kept arriving at the same conclusion: the typical asset-management playbook doesn't produce the alpha it claims to. The data is global, multi-decade, and unambiguous. So we built a different one. The equity-trading arm of FiduVentures is IMG Trading Systems. The goal is the same in any market — make the trade work — and we pursue it through the IMG Way: three deliberate departures from how the rest of the industry operates.
S&P Dow Jones Indices has published the SPIVA scorecard for more than two decades — a global comparison of active funds against their benchmarks. The headline is consistent across geographies and asset classes. The longer the time horizon, the more active funds fall behind.
The chart above is the surface of something structural. Across four decades of academic work and live practice, six forces have been shown to compress the alpha of traditional active management.
Same $1,000, same period, identical market exposure. Trading costs plus fees compound to roughly 19% of terminal wealth at typical institutional expense ratios — real AQR execution data across 21 markets confirms the cost scales steeply with position size.
Frazzini, Israel & Moskowitz 2018 · ICI 2024
About one in five active fund dollars sits in portfolios that hug the benchmark closely enough to be statistically indistinguishable from the index — while charging active fees. The 20-year academic review confirms this remains a structural problem.
Cremers, Fulkerson & Riley 2019 · ESMA 2020
Past returns attract flows. Flows grow AUM. AUM earns fees. Fees incentivise herding — which makes the manager's next return track the consensus. Younger managers with the most to lose from being wrong systematically herd toward their peers' portfolios to protect their careers.
Del Guercio & Reuter 2014 · Pástor, Stambaugh & Taylor 2015
Of every US domestic equity fund in the top quartile in December 2020, none remained there by December 2024. Past performance predicts future performance about as well as a coin flip — and the modest persistence that does exist is mostly momentum and fees, not skill.
SPIVA Persistence Scorecard 2024 · Fama & French 2010
What was once a research edge is now table stakes. Algorithmic trading has compressed the price-discovery window from minutes to milliseconds; the IMF documents AI further accelerating that trend. The information advantage of "knowing the company better" has been systematically arbitraged away.
Brogaard, Hendershott & Riordan 2014 · IMF Global Financial Stability Report 2024
On top of all this, manager-level decisions under uncertainty are subject to the same prospect-theory distortions. Funds managed by more loss-averse managers take on less downside risk, earn lower risk-adjusted returns, and are dramatically more likely to be terminated (Bodnaruk & Simonov 2016). Prospect theory predicts these distortions propagate into the asset prices themselves (Barberis, Jin & Wang 2021). Behaviour is the friction.
Bodnaruk & Simonov 2016 · Barberis, Jin & Wang 2021
To produce absolute returns in any market, IMG Trading Systems is built around three deliberate departures from how most asset managers operate. Each one is a small loss of orthodoxy. Together they are the thesis — and together they address a problem most active funds never solve: alpha that decays as the fund grows. Capacity is a function of how long you hold. Most active managers hold long enough for AUM to crowd out alpha; systematic evidence shows active skill exists but decays with scale (Pástor, Stambaugh & Taylor 2015). Our average holding is days, not quarters. The capacity ceiling is correspondingly higher.
Screens, signal validation, and sizing run through code. Humans run the parts where judgement adds value — fundamental sanity-checks, exception handling, risk overrides — but the moment of entry itself is signal-driven. The reason this matters is not a generic appeal to "bias"; it is that discretion at the entry point is precisely where active management systematically fails. At the manager level, the evidence is unambiguous: roughly one in five active funds hugs the benchmark so closely they are statistically indistinguishable from the index (Cremers, Fulkerson & Riley 2019); younger managers systematically herd toward their peers' portfolios to protect their careers (Del Guercio & Reuter 2014); and on aggregate, active management has not outperformed passive over multi-decade horizons (Pástor, Stambaugh & Taylor 2017). Signal-driven entry removes the discretion that drives these failure modes. The signal logic is documented, the review system writes down what worked and what didn't, and the process can be inspected, audited, and improved. That is not consistency — it is accountability.
Cremers, Fulkerson & Riley 2019 · Del Guercio & Reuter 2014 · Pástor, Stambaugh & Taylor 2017
Closing intraday is not a forecast — it is a structural risk-budget decision. The overnight return distribution is fundamentally different from the intraday one: it carries fatter tails and negative skew because the investor base during the overnight window is structurally different from the daytime one — household investors, foreign markets reacting to news, after-hours earnings flow (Lou, Polk & Skouras 2019). After-hours earnings announcements cause a price jump in over 90% of cases (Christensen, Timmermann & Veliyev 2025). The cleanest hedge is not to hold through the gap. Every intraday closure resets the gap-risk clock to zero.
Christensen, Timmermann & Veliyev 2025 · Lou, Polk & Skouras 2019
Most managers wait for the story. We enter on the chart — and that is where our rules stop and our judgement starts. Timing matters more than catching the bottom. Returns to momentum signals concentrate in the intermediate horizon, not the most recent month (Novy-Marx 2012): the tradable signal is the move that preceded the recent reversal, and the trader who waits for confirmation has already given up the bulk of it. Even published anomaly portfolios lose roughly 58% of their in-sample return after academic publication, because markets learn and consensus catches up (McLean & Pontiff 2016). The window belongs to whoever acts on the signal first — and the question of whether the position earns its overnight is ours.
Novy-Marx 2012 · McLean & Pontiff 2016
Roughly seven in ten positions we take are opened and closed within the same trading session. The remainder carry overnight — held only if the position has built enough buffer to absorb the additional overnight risk. No buffer, no carry. This is our version of margin-of-safety discipline: overnight carry is conditional — a position extends past the session only when it has built a buffer sufficient to absorb the gap risk it would otherwise inherit. Sizing is tiered by holding period — intraday largest, multi-day smaller, multi-week smaller still — because volatility scales with the square root of time, and overnight exposure adds a jump-risk component that does not diversify away.
Rules-based and repeatable. Each step has a defined input and a defined output. If a candidate doesn't pass a step, it doesn't advance.
We screen for in-play stocks — names with identifiable catalysts or names flagged by our premarket research — across the most liquid US and Hong Kong equities. Patterns flagged include accumulation, momentum shifts, volume anomalies, and multi-timeframe signal convergence. We generally trade no more than ten positions a day. Anything beyond that threshold is re-evaluated the next session.
Fundamental validation confirms the business behind the move. The trigger is confirmed on our internal charting stack with proprietary indicators tuned to the way we trade — not the off-the-shelf oscillators available on retail platforms. The proprietary backtester cross-checks the signal against historical analogues under realistic slippage.
Sizing is driven by signal strength, conviction, holding-period target, and risk parameters. Because volatility scales with √time, intraday positions can be sized larger than multi-day positions for the same risk budget.
Every position is monitored against its entry thesis in real time. If the technical signal fades or the pattern breaks, the position exits. Capital is recycled into the next signal.
After every session — and after every closed position regardless of outcome — our proprietary trade review system runs a structured analysis: what the trigger was, how the trade played, what we'd do differently, where the model was right and wrong. Findings are written, not narrated. Recurring patterns feed back into the model. Adjustments are made when evidence supports them, not when emotion demands them.
We own everything that decides when and how we trade: the indicators, the charting system, the backtester, the alerts, the review loop. Order routing and clearing sit with our broker — standard third-party infrastructure. The combination is what matters: our proprietary signal stack running on standard broker execution.
Oscillators and pattern detectors tuned to the specific triggers that show up on US and HK liquid names. Written for the signals we care about.
A charting system built around our indicators — not the other way around. Confluence across timeframes and asset classes is first-class.
Runs against historical tick data with realistic slippage and commission modelling. The same engine that runs the backtest runs the live book — so live performance is directly comparable.
Real-time alerts on signal triggers across the universe, routed directly to the desk via our own infrastructure. No third-party notification layer between the screen and the desk.
Every position gets a structured review post-close — trigger, outcome, deviation, adjustment. The system identifies recurring patterns and feeds them back into the model.
Order routing, clearing, and settlement sit with our broker. Standard third-party infrastructure — and where our proprietary stack runs.
Liquidity is not a filter we apply at the end of the trade decision — it is the strategy. Names without a tight spread and a deep book do not enter the universe. Risk management at the input layer, not the output layer.
IMG Trading Systems currently manages proprietary capital and is not actively seeking outside investors. When we're ready to consider aligned partners — institutional allocators, family offices, or strategic counterparties — we'll publish that here. If you'd like to be in touch when that opens up, send a brief note describing who you are and what you'd bring. We read everything; we respond selectively.