The Liquid Twenty: Stop Picking Stocks. Own the Busiest Instead.

One ranking rule. Twenty stocks. Traded only when a holding falls out of the buffer. How you hold India’s most-traded NSE names, weighted equally, without making a single analyst call — and what we can and cannot yet tell you about how it performs.

How often you re-run that ranking is a separate, practical choice. We currently check every trading day. Checking more often does not mean trading more often — the rank-25 buffer decides that, and on most days nothing crosses it.

§01 The One Number That Runs the Whole Strategy

Every stock in the Nifty 500 universe is ranked by a single number called Average Daily Value (ADV) — the average of (Closing Price × Volume) over the past 20 trading days. No fundamentals, no momentum scores, no analyst ratings. Just this number.

ADV answers one question: which stocks are large institutions actually trading in size right now? A high-ADV stock has deep liquidity, active price discovery, and serious institutional participation. It doesn’t gap 15% on thin air — when it moves, real money moved it.

ADV  =  Mean( Close × Volume )  over the past 20 sessions

Example (illustrative — India's largest private bank):
  Close ≈ ₹1,700   Volume ≈ 1.2 cr shares/day
  ADV   = ₹1,700 × 1,20,00,000 = ₹2,040 crore/day

This number is recalculated fresh at every rebalance. It is a rolling signal, not a stale annual average.

Once every stock is ranked, you take the top 20 — those with the highest ADV — and invest an equal rupee amount in each. Nothing else is evaluated. The ranking is the strategy.

§02 The Strategy in Eight Steps

Run through these steps each time you re-rank. Most times you will reach step 6, find nothing has crossed the line, and stop — checking is not trading.

1. Pull the universe. Start with all stocks in the Nifty 500 index. In our most recent run this resolved to 498 tradable NSE symbols, of which 319 carried enough clean price and volume history to be ranked.
2. Calculate ADV for each stock. For each name, compute the mean of (daily closing price × daily volume) over the 20 most recent trading sessions. This is one arithmetic operation per stock.
3. Rank by ADV, descending. The stock with the highest rupee turnover sits at rank 1. Sort the whole eligible universe by this number.
4. Identify the target portfolio. Your target is the top 20 stocks by ADV. Write them down.
5. Check existing positions. Compare your current holdings to the new top-20 list. Stocks that appear in both stay — no action needed.
6. Apply the buffer. Only sell a holding if it has dropped past rank 25. A name at rank 23 is not in the top 20, but it stays. This buffer is what stops a daily check from becoming daily trading — without it, a stock oscillating around the rank-20 line would be bought and sold repeatedly.
7. Execute the changes, if any. Sell the stocks that fell past rank 25. Buy the ones that entered the top 20. These are independent events — a name can enter without another leaving, which is why the holding count floats between roughly 19 and 21.
8. Do nothing, most of the time. The top 20 by turnover is a stable list; large-cap liquidity leaders do not reshuffle often. If nothing crossed rank 25 and nothing new entered the top 20, the correct action is none.

§03 What We Can and Cannot Show You

This is the part of a strategy article where you normally get a table of annual returns. We are not going to give you one, and it is worth being precise about why.

What we do have is the rule, stated completely above, and a live implementation that began recording on 17 August 2026. Every rebalance from that date forward is logged: which stocks entered, which left, what the basket weighed, and what it did next. That record is short — as of writing it is a few sessions old — and a few sessions of anything tells you nothing.

So we publish it as it accumulates, and we will not quote a return figure until there are enough completed rebalance periods for one to mean something. This is the same standard we hold our other screeners to: the four scanner modules on FutureGain each ship with an empty backtest field and a note saying so, and none of them displays a hit rate until at least 30 resolved signals sit behind it.

If you want performance figures before then, the honest path is to compute them yourself. The rule is fully specified — universe, ranking metric, basket size, buffer, cadence — with nothing held back. Any competent backtest over NSE data will reproduce it exactly, and you will trust your own numbers more than ours anyway.

§04 The Rebalance in Practice

Most periods, the top 20 barely changes. Large-cap liquidity leaders are structurally sticky — they stay at the top because their ADV is driven by index flows, options hedging, and institutional program trading, not transient retail interest.

When a swap does happen, it typically involves a stock that has had a liquidity event — a large block deal, a new options series, or inclusion in a derivatives basket — temporarily lifting its ADV above a long-standing member. The rank-25 buffer means you only act when the change is durable, not when a single high-volume day distorts the ranking.

The table below is a constructed example showing how the buffer resolves a rebalance decision. For the real current basket, see the live Liquid Twenty page.

Illustrative only — these stocks, ranks and ADV figures are invented to demonstrate the buffer rule, not drawn from any live scan. The rank-25 buffer means HINDUNILVR at rank 27 triggers a sell; a stock at rank 23 would be kept.

§05 Why Equal Weight Works

Most Indian investors default to market-cap weighting — owning more of the companies that are already the most expensive by market value. Equal weighting is the opposite bet: it systematically overweights the less-expensive large-caps and trims the extended ones at every rebalance.

In principle, this mechanical rebalancing acts as a disciplined “buy low, sell high” loop within the universe. When a stock rallies hard, the rebalance trims it; when one lags, the rebalance adds to it. No judgment required. Whether that mechanism actually pays for its costs in Indian large-caps is an empirical question this article does not claim to have settled.

The liquidity filter adds a second layer: by restricting the universe to only the most-traded names, the strategy avoids the illiquidity premium entirely. Every position can be entered and exited at tight bid-ask spreads on any trading day.

§06 Realistic Costs and Taxes

Whatever this strategy returns gross, that is not what you keep. There are two unavoidable cost buckets: transaction costs on every swap, and STCG tax on every profitable exit — holding periods here run to months at most, so every gain is short-term.

Cost depends on how often a name actually crosses rank 25 — a property of the market, not of how often you look. Because we are not publishing a return figure, the drag is best expressed as a formula you apply to whatever gross number your own backtest produces, and to whatever swap count your own data shows:

Swaps per year        = S   (count them in your own backtest —
                             we do not have enough live history yet)
Round-trip cost       = 0.2%  (0.1% per leg, incl. impact)
Each swap touches     = 1/20 of the book

Annual transaction drag ≈ S x 0.2% x (1/20)
                        ≈ 0.01% per swap

  S = 12  →  ~0.12%      S = 40  →  ~0.40%
  S = 24  →  ~0.24%      S = 80  →  ~0.80%

Then: net = (gross - transaction drag) x (1 - 0.20 STCG on the gain)

Worked through, the tax alone removes roughly a fifth of whatever you make. On a hypothetical 12% gross year that is about 2.3 percentage points; on a hypothetical 20% year, closer to 3.9. Tax drag shrinks if the position sits inside a tax-advantaged structure, or if losses from down periods can be carried forward against the gains.

§07 Honest Limitations

§08 Run It on Live Data

We run this scan against the Nifty 500 and publish the result. The Liquid Twenty page shows the current twenty holdings with their turnover ranks, the five names sitting in the rank-21-to-25 buffer, and the sector mix — plus a sizing tool that converts the basket into exact share counts for whatever capital you are working with, rather than the fixed ₹1 lakh the scan assumes.