What actually happens between rebalancing dates
I set my factor targets at the start of a period and felt confident. Then the market
moved, and by the next rebalancing window my actual exposures had drifted in directions
I had not anticipated. I could only trade part of the way back because of turnover
limits. The residual tilt carried into the next period, where it interacted with new
drift in ways my original plan had not accounted for at all. That is the experience that
makes multi-period factor targeting worth taking seriously. The frameworks on this site
exist to address exactly that sequence of events — not by eliminating the difficulty,
but by building it into the planning logic from the start rather than discovering it as
a recurring surprise. Market drift is continuous. Turnover constraints bind in every
period. And the interaction between the two accumulates across rebalancing cycles in
ways that single-period models are simply not designed to track. What follows on this
page works through the core mechanics behind that problem: how factor exposures drift,
why turnover constraints make drift management path-dependent, and what it means to plan
across multiple rebalancing intervals rather than optimizing each one in isolation.
Nothing here constitutes financial advice of any kind. Past performance of any
analytical framework does not guarantee future results.
Drift is not a measurement error or an execution failure — it is the structural result of
holding positions in a market that moves continuously between rebalancing dates, and it
compounds when turnover constraints prevent full correction in any single period
How factor exposures drift and why the standard response falls short across rebalancing periods
Understanding drift requires separating two things that often get conflated: the target
exposure and the achievable exposure given the constraints in place.
A factor target describes where the analyst wants the portfolio's exposures to sit. Market drift describes the continuous movement of actual exposures away from that target as prices change between rebalancing events. The gap between the two is not static. It grows when market conditions are volatile and shrinks when they are calm, but it rarely stays at zero for long without active intervention.
The standard response to drift is straightforward: at each rebalancing date, calculate the
trade that closes the gap between actual and target exposure, subject to turnover limits and
transaction costs, and execute. That approach works well when drift is small and turnover
capacity is generous. When those conditions do not hold, its limitations become visible
quickly.
When turnover constraints bind — when the gap between actual and target is larger than the available trading budget can close in a single period — a residual exposure carries forward into the next period. That residual then interacts with the drift that accumulates during the next interval, producing a starting point at the following rebalancing date that the previous period's trade was not designed to handle.
Over several rebalancing cycles, this produces a factor trajectory that is reactive rather than intentional. Each period's trade is locally optimal given the information available at that rebalancing date, but the sequence of trades across periods does not form a coherent plan. The result is a portfolio whose actual factor exposures oscillate around the target in ways that no single period's analysis fully explains.
Path dependency explained plainly
Path dependency is the property that makes multi-period factor management different in kind,
not just degree, from single-period optimization. The trade executed today does not only
affect today's factor exposures. It determines the starting point for the next
rebalancing, which shapes the trade available then, which affects the period after that.
Each decision in the sequence constrains the decisions that follow it, and that constraint
cannot be managed by looking only one period ahead at each rebalancing date. When turnover
limits are tight, the path-dependent nature of the problem becomes the dominant
consideration. A trade that looks slightly inefficient in a current-period view can be
clearly correct when its effect on the next two or three starting points is made visible.
Multi-period frameworks are built around that insight.
The practical consequence of ignoring path dependency is that turnover budgets get used reactively rather than strategically. A single-period optimizer will typically consume the full available budget chasing the current-period target gap, because that is what minimizes current-period tracking error. A multi-period framework will sometimes recommend a smaller trade this period in order to preserve turnover capacity for a larger or more important correction in the following interval. That difference in logic produces measurably different factor trajectories over a full rebalancing cycle, particularly when market conditions are generating persistent drift in a consistent direction.
The difference is not in the complexity of the mathematics but in the question being asked — multi-period frameworks ask about the full planning horizon from the start, not one period at a time
What multi-period factor frameworks actually do differently from sequential single-period approaches across a rebalancing cycle
The clearest way to see the difference is to follow the same rebalancing decision through both lenses and observe where the logic diverges.
In a single-period framework, the analyst arrives at each rebalancing date, observes
the current gap between actual and target factor exposures, and calculates the trade
that closes as much of that gap as possible within the turnover constraint. The
decision is complete. The next period starts fresh with the same logic applied to
whatever starting point the previous trade produced.
In a multi-period framework, the analyst arrives at the same rebalancing date with an additional layer of information: a view of how market drift is likely to move exposures over the next one or two intervals, what turnover capacity will be available at future rebalancing dates, and what starting exposures those future decisions will inherit from the current trade. The current-period trade is chosen with all of that in view.
The result is not necessarily a larger or more aggressive trade. It is often a more measured one — a trade that closes a meaningful portion of the current gap while preserving options for the periods ahead. When drift has been moving exposures in a consistent direction, a multi-period framework will sometimes recommend beginning to position against the anticipated future drift before it has fully materialized, using part of the current-period turnover budget proactively.
Evaluating whether a multi-period approach is working requires metrics that look across the full planning horizon rather than at a single rebalancing snapshot. A framework that produces tighter average tracking error across six periods is more valuable than one that minimizes tracking error at each individual rebalancing date but allows it to widen significantly in the intervals between. That distinction is easy to state but requires deliberate measurement to observe in practice.
When turnover limits are treated as costs to minimize at each rebalancing event, they remain reactive constraints. When they are treated as planning parameters with a budget to allocate across periods, they become a tool for managing factor trajectories more deliberately over time
The framing of turnover constraints determines how they are used and what kind of factor trajectory results from that use.
Turnover constraints as planning parameters rather than execution costs across the full rebalancing cycle
The framing of turnover constraints determines how they are used and what kind of
factor trajectory results from that use.
- A common approach treats turnover limits as a cap on the size of the trade that can be executed at each rebalancing date. Within that cap, the optimizer finds the trade that minimizes the gap between actual and target factor exposures. The limit is a constraint in the technical sense: a boundary that the solution must not cross.
- A multi-period approach treats the same turnover limit differently. Rather than asking what is the largest gap I can close this period within the limit, it asks how should I allocate my total turnover capacity across the next several rebalancing intervals to produce the most consistent factor trajectory over the full planning horizon. The limit becomes a budget with a time dimension rather than a per-period ceiling.
- That reframing changes the trades that get recommended. When the multi-period framework anticipates that market drift will generate a larger-than-usual exposure gap in two periods' time, it may recommend using less turnover capacity this period in order to have more available when the larger correction is needed. A single-period approach has no mechanism for that kind of forward-looking allocation because it resets its assumptions at each rebalancing date.
Three concepts at the core of multi-period factor exposure management
Every framework discussed on this site is built on the interaction among three core mechanics. Understanding each one separately makes the multi-period planning logic considerably more concrete and easier to reason about in practice when working through specific rebalancing scenarios.
Market drift and its cumulative effect
Between any two rebalancing dates, market price movements continuously shift the portfolio's actual factor loadings. When those shifts are small and rebalancing is frequent, drift is easy to manage. When turnover is constrained and rebalancing windows are wider, drift accumulates in ways that interact with the residuals left by previous incomplete corrections. Recognizing drift as cumulative and directional rather than random and self-correcting is the first step in building a framework that addresses it honestly.
Turnover budgeting across the planning horizon
Turnover limits are not simply per-period execution caps. When viewed across a planning horizon, they represent a total allocation of trading capacity that must be distributed across multiple rebalancing intervals. How that capacity is distributed — how much is used this period versus reserved for anticipated future corrections — determines the shape of the resulting factor trajectory as much as the choice of target exposures does. Multi-period frameworks make that allocation explicit and deliberate rather than leaving it as an implicit consequence of sequential single-period decisions.
Path dependency and its planning implications
Because today's trade becomes tomorrow's starting exposure, and tomorrow's starting exposure shapes the trade available then, each rebalancing decision is connected to all subsequent ones through the portfolio's state. That connection is path-dependent: the optimal sequence of trades is not simply the concatenation of individually optimal trades. When turnover constraints bind, recognizing and accounting for path dependency is what separates a coherent multi-period factor framework from a sequence of reactive single-period adjustments that happen to follow one another in time.
Key concept
Measuring consistency across the full rebalancing horizon
Measuring whether a factor framework is delivering consistent exposures over
time requires looking across periods rather than evaluating accuracy at each
individual rebalancing snapshot taken in isolation.
Useful consistency metrics track average tracking error across the full planning horizon, the variance of that tracking error across periods, and whether residual exposure gaps show persistent directional bias that indicates the framework is systematically undershooting or overshooting its targets.
Five practical questions that multi-period frameworks are built to answer
These are the questions that come up most often when thinking carefully about factor
exposure management under real-world constraints. Each one points directly to a
limitation of single-period approaches and illustrates why a multi-period planning
horizon produces more useful and coherent rebalancing decisions over time across a full
cycle.
-
01
How much of the current gap should be closed now versus spread across future periods?
This depends on how much of the gap is likely to be offset by anticipated drift in the next interval, what turnover capacity will be available at future rebalancing dates, and whether the full trading budget is more valuable now or later given the expected market conditions across the planning horizon.
-
02
When should turnover capacity be held in reserve rather than fully deployed?
When drift is moving exposures in a consistent direction and that trend is likely to continue, using the full current-period turnover budget on the current gap may leave insufficient capacity to address a larger gap at the next rebalancing date. Multi-period frameworks identify those situations in advance and allocate capacity accordingly across the full horizon.
-
03
How does the reachable factor exposure set shrink under binding turnover constraints?
Not every factor trajectory is achievable when turnover is limited. The set of exposures reachable within a given planning horizon is narrower than a single-period model suggests. Understanding that reachable set explicitly allows the analyst to evaluate which target trajectories are practically feasible and which require either relaxed constraints or a longer planning horizon to achieve.
-
04
How should residual tilt from previous periods be incorporated into the current rebalancing decision?
Residual tilts are not simply errors to be corrected as quickly as possible. They represent committed positions with their own factor implications across the remaining planning horizon. A multi-period framework treats residuals as forward-looking planning inputs rather than backward-looking correction targets that must be eliminated at the earliest available rebalancing date.
-
05
What metrics reveal whether a factor framework is maintaining consistency across periods rather than just accuracy at each rebalancing date?
Single-period accuracy metrics — tracking error at the rebalancing date, for example — do not capture how exposures behave in the intervals between rebalancing events. Cross-period consistency metrics that look at average and variance of tracking error across the full planning horizon provide a more complete and honest picture of how well a framework is performing over time.