Free lesson · Alternative data
Availability delays and signal decay
Open interactive lessonPractice calculationsExplore labs
Start with the idea
A measurement may be accurate yet arrive too late to improve a trading decision.
Symbols, units & horizon
- a₀: expected gross edge at reference availability, return fraction over the chosen holding horizon
- d≥0: additional latency in seconds
- h>0: assumed edge half-life in seconds
- a(d): remaining expected edge under an exponential decay assumption
When and why to use this
Compare vendor delivery and computation choices under realistic timing constraints.
A measurement may be accurate yet arrive too late to improve a trading decision.
Distinguish economic occurrence, public publication, vendor delivery, feature completion and order arrival. A stylized exponential decay curve makes the assumed cost of delay explicit. Its half-life must be estimated or stress-tested; it cannot be chosen after observing test returns.
Slow signals can remain useful when a fast reaction has passed, but this needs a defined horizon and benchmark. Information may arrive in bursts rather than decay smoothly, so compare several delay scenarios.
Availability delays and signal decay
- Express delay as a number of half-lives d/h.
- Each half-life multiplies edge by one half.
- Multiply the original edge by 2 raised to negative elapsed half-lives; this is an assumed model, not an identity of market behavior.
An initial edge of 20 basis points, half-life 60 seconds and delay 120 seconds imply .002×2^(−2)=.0005, or five basis points.
Apply it in a strategy
- Freeze inputs at the stated decision time and record their units.
- Compare vendor delivery and computation choices under realistic timing constraints.
- Recompute the example, then change the material assumption and explain the difference.
Research deliverable
Availability delays and signal decay: produce the worked calculation, a timestamped input record and a written decision addressing this limitation: Incorrect half-life or using revised publication timestamps can grossly overstate remaining edge.
These are synthetic mechanics examples, not historical performance or paper replications. Module evidence and research boundaries record the 12 September 2026 review.
Python implementation
Self-contained teaching example. Python 3.10+; dependencies and input conventions are shown in the code and notation. Run in your own Python environment.
# Python 3.10+; standard library and NumPy only.
# Synthetic teaching inputs; conventions and units are defined in the notation above.
def delayed_edge(initial_edge,delay_seconds,half_life_seconds):
if delay_seconds<0 or half_life_seconds<=0: raise ValueError('Invalid delay or half-life')
return initial_edge*2**(-delay_seconds/half_life_seconds)
assert delayed_edge(.002,120,60)==.0005
print(delayed_edge(.002,120,60))Continue learning
Alternative Data: Measurement, Text & Incremental Value — all lessons- From a sampled panel to a population estimate
- A reproducible dictionary score for text
- An event return needs a predeclared benchmark
- Availability delays and signal decay
- Noisy proxies and attenuation
- Measure improvement against a frozen baseline
- From forecast accuracy to a costed decision
- A data investment includes coverage, access and ongoing costs
Quantitative finance and development glossary · Python resources and libraries · Research sources and limitations