Concept

Wavelet Decomposition

Wavelet Decomposition is a Statistics concept.

What is wavelet decomposition?

Wavelet decomposition splits a time series into components at different scales: a smooth approximation capturing the slow trend, plus layers of detail capturing progressively faster fluctuations. Unlike FFT spectral analysis, which describes a series in terms of sines that extend over the whole sample, wavelets are localized little waves, so the decomposition says not just which frequencies are present but when. That time-frequency localization is the whole point for markets, where the character of price movement changes constantly.

The standard machinery is the discrete wavelet transform (DWT) and its shift-invariant variant (often called MODWT), built on wavelet families such as Haar or Daubechies. Each level of the transform halves the frequency band: level-1 detail holds the fastest wiggles, level-2 the next octave down, and so on, with the final approximation holding what remains. Summing all components reconstructs the original series exactly, which makes wavelets a principled form of signal decomposition rather than an ad hoc smoothing stack.

Traders and researchers use the decomposition to separate concerns: denoise a series by suppressing the finest details, extract a trend by keeping only the approximation, or study how volatility and correlation behave at different horizons. The recurring caveat is the boundary problem. Wavelet filters need data on both sides of a point, so values near the right edge of the sample are computed from padded or asymmetric data and revise as new bars arrive, a repainting behavior that has burned many an unwary backtester.

How it's calculated

There is no single formula; the canonical scheme is the pyramid algorithm of the discrete wavelet transform. One level of decomposition is:

a_j[k] = sum_n h[n - 2k] * a_(j-1)[n]
d_j[k] = sum_n g[n - 2k] * a_(j-1)[n]
x = A_J + D_J + D_(J-1) + ... + D_1
a_(j-1): approximation coefficients entering level j (a_0 is the raw series)
h: low-pass filter of the chosen wavelet family (e.g. Haar, Daubechies-4)
g: matching high-pass filter
a_j, d_j: approximation and detail coefficients at level j
A_J, D_j: reconstructed approximation and detail series; their sum returns the original series x

The MODWT variant skips the downsampling by 2, keeping every level aligned with the original bars, which many time-series practitioners prefer.

Wavelet choice and depth J are tuning decisions; Haar is the simplest and most edge-friendly, longer wavelets are smoother but worsen boundary effects.

How traders use it

  • Denoising: suppress or threshold the finest detail levels and reconstruct, producing a smoothed series that adapts locally instead of applying one fixed bandwidth everywhere; the aggressiveness is set by how many levels are removed.
  • Multi-horizon analysis: examine detail energies by level to see which timescales carry the movement, a quantitative cousin of looking at multiple chart timeframes at once.
  • Feature construction: wavelet coefficients or per-level variances serve as model inputs, letting a classifier see short-term and long-term behavior as separate features rather than one entangled series.
  • Trend extraction: keep the approximation and coarsest details as a trend estimate, an alternative to tools like the Hodrick-Prescott filter with more explicit control over which scales survive.
  • The critical limitation for live trading: near the current bar the transform relies on boundary handling, so recent values revise as data arrives. Any strategy built on wavelet output needs the discipline of repaint-safe engineering, or a causal reformulation.

Wavelet decomposition vs related concepts

FFT Spectral Analysis: Fourier analysis assumes the same frequency mix holds across the whole sample; wavelets localize in time as well as frequency, which suits markets whose rhythm changes from month to month.

Signal Decomposition: The umbrella idea of splitting price into trend, cycle, and noise components; wavelet decomposition is one specific, exactly invertible way to do it, with scales fixed by octave rather than fitted.

Kalman Filter: A Kalman filter extracts components causally, updating from past data only, so it never repaints; wavelet transforms look both directions in time, giving cleaner in-sample separation but unreliable values at the live edge.

Concept family

Statistics

46 concepts mapped · 46 in the Library

Wavelet Decomposition FAQ

What is the advantage of wavelets over Fourier analysis for price data?

Localization. Fourier describes the whole sample with global sinusoids, so a burst of volatility in one month smears across the entire spectrum. Wavelets pin each component to a time and a scale.

Do wavelet-based indicators repaint?

Values near the most recent bars generally revise as new data arrives, because the transform needs future observations that do not exist yet at the edge. Backtests that ignore this can look far better than live results.

Which wavelet family should I use?

Haar is simple, causal-friendly, and handles jumps well; Daubechies wavelets of modest length are the common default for smoother decompositions. In many applications the choice of decomposition depth matters more than the family.

Is wavelet denoising better than a moving average?

It is more flexible: it can remove fast noise while leaving sharp legitimate moves largely intact, where a moving average blurs everything equally. The price is added complexity and the edge-effect problem.

Build Wavelet Decomposition your way.

Quant writes, tests, and refines it with you — then it runs on LuxAlgo charting or ports to TradingView.