Process before performance

Summary: Two fund managers. Same market, same data. On any given day, the results look similar. Over a decade, the divergence is structural. And it traces back to something most investors never think to ask about before they invest.

Every recipe exists for the same reason: consistent output. Not because the cook lacks talent, but because talent alone does not scale. The same logic applies to heart surgery, aircraft maintenance and fund management also. Without process, you are not investing. You are improvising.

In an AMC, process is required across every function: compliance, operations, risk, fund accounting. But it is most consequential in investment and research, where capital allocation decisions originate. That is where investors feel the difference between discipline and discretion.

Why process is the upstream guarantee

An investment process is the structured, repeatable steps that take an AMC from data to research to portfolio decisions. It is also the definition of fiduciary duty: acting in investor’s best interest, not fund manager’s convenience. Most investors measure fiduciary care through returns. That’s the wrong meter. Returns are an output. Process is the input that determines their quality and consistency.

Consider two fund managers: same market, same data. One builds a portfolio based on a defined rulebook. The other, on gut-feeling and discretion. On any day, results look similar. Over a decade, the divergence is structural. Process removes the two biggest threats to investment outcomes: human bias and inconsistency. Behavioural biases like overconfidence, recency, and herding erode returns in discretionary frameworks because even experienced investors are susceptible to them. A defined process designs them out rather than managing them.

NJ AMC’s six-step investment process

Step 1: Data validation, verification and cleansing. Data is the raw material. Its quality determines everything downstream. NJ AMC subjects third-party data to a three-stage audit: validate the source, verify consistency across providers and cleanse for errors and outliers.

Step 2: Development of factor parameters and hygiene check. Clean data feeds into factor parameters, the lenses for evaluating stocks.

NJ AMC customises metrics like ROCE, ROE and free cash flow rather than using off-the-shelf values. Even a standard measure like ROE can be computed multiple ways, each producing different backtests. Every parameter undergoes a hygiene check before entering NJ AMC’s Parameter Library.

Step 3: Parameter robustness testing. While new ideas lead to new parameters, not all earn a place in portfolio construction. A standardised robustness testing process separates the strong from the weak. The logic is straightforward. If a parameter is genuinely robust, the stocks it ranks highest should outperform the middle-ranked stocks, which should outperform the worst-ranked, in both return and risk. To test this, the universe is divided into equal slices holding stocks of similar rank. A robust parameter behaves predictably: the top slice beats the middle, the middle beats the bottom and volatility rises as rank falls. NJ AMC’s SmartBeta Research Platform applies this test across more than 1,200 companies and 20 years of data. Only parameters that pass are promoted for portfolio construction.

Step 4: Idea generation and portfolio construction. With validated parameters, the research team builds not one but thousands of portfolio variants across four key construction decisions: universe selection, stock selection, portfolio allocation methodology and rebalancing frequency. Quantitative creativity operates here, but within structured, testable discipline.

Step 5: Analysis of the portfolio strategy. Every model portfolio is subjected to three layers of scrutiny across nearly 20 years of data spanning multiple market cycles. Performance analysis covers metrics such as CAGR, Sharpe ratio, maximum drawdown and rolling returns, examined over overlapping and non-overlapping periods. Attribution analysis identifies where returns came from. Churn analysis measures portfolio turnover because high churn raises transaction costs and erodes strong strategies.

Step 6: Model finalisation and implementation. The research team presents findings to an independent Investment Committee, which evaluates research rigour, risk-return potential, implementation feasibility and model suitability. Only strategies that clear this gate are approved for deployment. A pre-deployment liquidity analysis ensures execution at scale without meaningful market impact.

What this means for the investor

Process is not an internal operational footnote. It is a promise. It means decisions are data-driven rather than emotional, consistent across teams, true to label through market cycles and transparent enough to be audited and explained.

The edge that scales

By removing human bias and subjecting every decision to a structured, repeatable framework, NJ AMC fulfils its fiduciary commitment through process, because ultimately, the long-term success of any investment strategy is directly proportional to the strength and resilience of the investment process.

Nirmay Choksi is the Executive Director - Research at NJ Asset Management Private Limited. The views expressed above are his own.

Also read:  ‘Investing is a marathon, not a sprint’