Several symbols can carry one shared risk
Correlation describes how two measured series move together within a sample. For stock research, calculate it from returns rather than raw price levels, which can trend together and produce misleading results. A value near +1 means closely aligned movement in that sample, near −1 means opposing movement, and near zero means little measured linear relationship. None of those values is a promise about the next market shock.
This matters particularly for a grid strategy that adds positions as prices move against it. Three technology stocks can all trigger additional buy levels during the same sector decline. Each grid may appear small in isolation while the account's combined exposure grows quickly. FINRA's concentration guidance emphasizes looking through investments to shared sectors and underlying holdings. Counting symbols is an inadequate substitute for measuring the capital at risk.
Calculate comparable return samples
A simple daily return is today's adjusted closing price divided by the prior close, minus one. Align dates and use the same currency and adjustment convention. The sample correlation is the covariance of the two return series divided by the product of their standard deviations. Spreadsheet functions can calculate it, but they cannot decide whether your inputs answer the intended question.
Compare several preselected windows, such as a shorter and longer history, and examine drawdown episodes separately. Do not choose the least-correlated window after seeing the result. Missing observations, stale prices, splits and asynchronous market closes can distort the output. A matrix should show the sample dates and number of observations; a correlation without those details is difficult to evaluate or reproduce.
A worked basket exposure example
Imagine three unleveraged hypothetical stock grids. Each begins with $1,000 of exposure and can add two more $1,000 levels. Initial combined exposure is $3,000, but maximum planned exposure is $9,000. On a $10,000 account, that is 90% of equity before price changes. If all grids fill their levels and the basket subsequently falls 20% from its average entry values, the approximate mark-to-market loss is $1,800, or 18% of starting equity, before costs.
The calculation is an exposure scenario, not a predicted loss or a maximum guarantee. A leveraged CFD account can face financing charges and margin liquidation, and its cash deposit does not equal its notional exposure. A stock can gap below a planned basket exit, suspend trading or fail permanently. Adaptive spacing changes entry distances; it does not remove a shared economic exposure or guarantee mean reversion.
| Stage | Exposure per symbol | Three-symbol exposure | Share of $10,000 equity |
|---|---|---|---|
| First level | $1,000 | $3,000 | 30% |
| Second level | $2,000 | $6,000 | 60% |
| Third level | $3,000 | $9,000 | 90% |
Stress shared factors, not just pairwise numbers
A low historical correlation can disappear when companies face a common financing or demand shock. Identify sectors, customer groups, interest-rate sensitivity, currencies and index membership. Include holdings outside the grid account. If the same company appears in an ETF and as a grid position, its effective weight is larger than the grid statement alone shows. Diversification can reduce some risks while leaving broad market exposure intact.
Try scenarios in which all symbols fall together, spreads widen and one position cannot be closed promptly. Estimate equity loss at every filled level, required margin where relevant and the cash needed for charges. A scenario in which only one stock declines may be useful, but should not be the sole test. Also test what happens when the account reaches its exposure ceiling and cannot add another recovery order.
Basket review checklist
Keep the stress worksheet with the settings used for each test. A positive historical basket return does not demonstrate that the grid can survive every longer trend. Record the worst floating loss and duration as well as closed trades. Changing position sizes after inspecting a favorable sample should trigger a fresh independent evaluation rather than a stronger claim of safety.
- Document the maximum levels, size at each level and total notional exposure across every strategy instance.
- Use aligned adjusted returns and label the correlation period, observations and calculation method.
- Aggregate sectors and shared holdings, including direct shares, ETFs and other accounts.
- Test simultaneous declines, gaps, financing costs and unavailable exits without assuming perfect stop fills.
- Define a basket loss rule and exposure ceiling before adding positions; review the method on independent data.
Frequently asked questions
Does low correlation make a grid safe? No. Correlations change, losses can be permanent and the grid's growing exposure still needs a hard limit.
Should correlation use prices or returns? Returns are generally the relevant input for measuring co-movement of investment changes. Specify the method and data adjustments.
Can adding more stocks reduce every loss? No. Extra positions can increase shared exposure, transaction costs and the difficulty of exiting during stress.
