Portfolio research

A good cycle is not yet a portfolio

Cycle backtests answer how individual rule-based entries behaved. A real portfolio adds position overlap, sector concentration, cash usage, trading costs, capacity, and the discipline required to hold or exit several names at once.

The difference between a cycle and a portfolio

A cycle usually begins when a ticker first meets an entry condition and ends at a defined exit. Backtest tables can therefore summarise thousands of independent historical examples. A portfolio cannot invest in every example at once. It has a finite number of holdings, a finite cash balance, and may receive several highly correlated signals during the same market phase.

This is why a strategy with an attractive average cycle return can still disappoint in practice. Its winners may cluster in one sector, its signals may arrive after a market has already advanced, or its drawdowns may overlap when risk appetite changes.

What a practical portfolio test should constrain

ConstraintWhy it matters
Maximum number of holdingsForces the test to choose among simultaneous candidates instead of assuming unlimited capital.
Position weightingEqual-weight and volatility-aware approaches can produce materially different exposure.
Sector capLimits the chance that several apparently separate picks are one commodity or industry bet.
Market-score gateTests whether fresh entries behave differently in Risk-on and Selective conditions.
Turnover and costsPrevents short holding periods from looking better than a tradable implementation would allow.
Capacity and liquidityFilters out trades that are hard to execute without changing the price.

How to read portfolio results

Look beyond total return. Compare maximum drawdown, time under water, turnover, cash exposure, number of active positions, and the distribution of returns across sectors and years. A robust portfolio result should not depend entirely on one exceptional period or a handful of illiquid names.

TrendRadar uses portfolio research to challenge promising cycle results. It is most valuable when it reveals that an appealing rule needs a sector cap, a market-regime filter, fewer concurrent positions, or a less aggressive entry threshold.

A cautious implementation mindset

  1. Start with a small, diversified paper or research portfolio rather than assuming historical capacity is available.
  2. Use an explicit maximum position size and sector exposure before any entry is considered.
  3. Record slippage, fills, taxes, and decisions that differ from the backtest rules.
  4. Review the gap between the original test and real execution instead of changing rules only after a loss.

Limitation

Portfolio simulations still rely on historical assumptions. They cannot guarantee future correlations, liquidity, or execution quality. Their role is to make a strategy harder to fool oneself with, not to turn it into a promise.