Syntetos–Boylan Approximation (SBA): The Bias-Corrected Intermittent Demand Forecast
Calculate SBA as a Croston bias correction, distinguish forecasts from safety stock and check the current LineNow method routing.
Jainul Vaghasia/Published /Updated /8 min read
Use the definition
Turn procurement terms into an operating system.
This reference page should help you understand the concept first. When the term affects purchasing execution, LineNow connects it to live POs, supplier replies, receiving, and accounting handoff.
Before a closed-loop procurement platform — a system where each buying step stays connected in one operating record, from demand signal to purchase order to supplier reply to receiving — can place an order, it needs a demand forecast. Different demand patterns warrant different forecast candidates. Simple averages or exponential smoothing can be useful baselines, including for intermittent demand; evaluate their out-of-sample errors rather than declaring one universally unsuitable. The Syntetos–Boylan Approximation (SBA) is the bias-corrected method for forecasting intermittent demand: the standard correction to Croston's original method, which has known bias under the assumptions studied in the research.
Quick answers
What is the Syntetos–Boylan Approximation? SBA is a modification of Croston's method for intermittent demand forecasting. It applies a deflating correction factor (1 − α/2) to the Croston forecast to remove a known positive bias, where α is the exponential smoothing constant used to update the demand-size and demand-interval estimates.
Why does SBA exist? Syntetos and Boylan (2001, 2005) proved that Croston's method has a systematic positive bias — it over-estimates mean demand, especially when the smoothing constant α is large. The inventory effect depends on the data, forecast error and ordering policy; it is not a guaranteed over-order on every item.
What is the SBA formula?
SBA forecast = (1 − α/2) × (d̂ / p̂)
where d̂ is the smoothed estimate of demand size (conditional on a non-zero period), p̂ is the smoothed estimate of the demand interval (average number of periods between non-zero demands), and is the smoothing constant (typically 0.05–0.20 for intermittent series).
Read before ordering
A dense operator briefing for teams that need sharper buying, cleaner supplier follow-up, and fewer expensive surprises.
α
When is SBA used instead of simple exponential smoothing? SBA is used for items classified as intermittent or lumpy in the SBC demand classification framework — items with an Average Demand Interval (ADI) above 1.32. Items with ADI ≤ 1.32 require a separately selected method; the classification itself does not mandate exponential smoothing.
The problem: why Croston's method over-estimates
Croston (1972) proposed a sensible approach to intermittent demand: rather than averaging over all periods — while retaining zero-demand periods in the timing information — separately smooth the demand size when it occurs and the interval between occurrences.
Define:
d_t = demand in period t (zero in zero-demand periods, positive otherwise)
q_t = the most recently observed non-zero demand size
p_t = the most recently observed inter-demand interval (number of periods since the previous non-zero demand)
This is intuitive: the expected demand rate is (average demand size) ÷ (average interval between demands). If an item sells 4 units every 5 days on average, the expected daily rate is 4 / 5 = 0.8 units/day.
The problem Syntetos and Boylan identified is a statistical artifact: the ratio of two separately smoothed estimators is not itself unbiased. Taking the statistical expectation of the ratio d̂/p̂ gives a value larger than the true mean demand rate μ. The bias is approximately proportional to α/(2−α), meaning larger smoothing constants amplify the over-estimation.
The approximation describes statistical bias under model assumptions. It does not establish a fixed over-order percentage, the right safety buffer or unchanged service for a particular catalog.
The approximation describes statistical bias under model assumptions. It does not establish a fixed over-order percentage, the right safety buffer or unchanged service for a particular catalog.
The SBA correction
Syntetos and Boylan's correction is compact:
SBA forecast = (1 − α/2) × (d̂ / p̂)
The factor (1 − α/2) deflates the Croston estimate back toward the true mean. At α = 0.10, the correction factor is 0.95 — a 5% reduction. At α = 0.20, it is 0.90 — a 10% reduction. This closely matches the bias it removes.
SBA is an approximate bias correction, not a guarantee of lower inventory or unchanged service. Syntetos and Boylan compared methods on 3,000 automotive intermittent-demand series; the paper reports better out-of-sample performance for their proposed method in that study. That result should not be restated as a universal 11% inventory saving. See the authors’ research record.
For the underlying separate smoothing of demand size and inter-arrival time, see Forecasting: Principles and Practice. Test the method on the business’s own data and decision horizon.
A worked example
A specialty retailer carries a slow-moving imported condiment — call it Item X. Here are its sales events over the last 30 days (days with sales shown; all other days are zero):
Day 3: 2 units · Day 6: 3 units · Day 9: 1 unit · Day 14: 2 units · Day 19: 2 units · Day 23: 1 unit · Day 27: 2 units · Day 30: 3 units
8 non-zero events in 30 days.
Step 1: Compute ADI.
ADI = 30 days / 8 non-zero periods = 3.75 days
ADI = 3.75 indicates intermittent arrivals. The non-zero sizes have sample variance 4/7 and mean 2, so CV² = (4/7)/4 ≈ 0.143. Together these put the example in the intermittent quadrant; SBA is a candidate to evaluate.
Croston base demand = 0.53224 × 7 ≈ 3.73 units
SBA base demand = 0.49232 × 7 ≈ 3.45 units
The example’s SBA forecast is 7.5% lower than its uncorrected Croston forecast. That does not make safety stock or the final PO exactly 7.5% lower: uncertainty estimates, inventory position, buffers and supplier packs also affect the order. Measure achieved service separately.
SBA in the SBC demand classification
SBA does not stand alone. It is one output of the Syntetos–Boylan–Croston (SBC) demand classification framework — the standard method for deciding which forecasting model to apply to each item. The framework classifies every SKU using two axes computed from historical demand:
ADI (Average Demand Interval): the average number of periods between non-zero demand events. ADI = total periods ÷ periods with non-zero demand.
CV² (coefficient of variation squared): the squared coefficient of variation of non-zero demand sizes. CV² = variance(non-zero demand) ÷ mean(non-zero demand)².
Demand pattern
ADI
CV²
Forecasting method
Smooth
≤ 1.32
≤ 0.49
Exponential smoothing
Intermittent
> 1.32
≤ 0.49
SBA
Erratic
≤ 1.32
> 0.49
Exponential smoothing + wider safety stock
Lumpy
> 1.32
> 0.49
SBA + manual adjustment
The 1.32 and 0.49 thresholds are commonly used classification conventions, not universal performance boundaries. The method column lists candidates; choose with validation on the relevant horizon.
SBA is a candidate for intermittent and lumpy demand, not a universally prescribed winner. Known events, declining demand and limited history may require a different method or an explicit manual scenario.
What this means for SMB operators
Most retail and restaurant operators carry a mix of demand patterns. A specialty retailer running 150 active SKUs might have 80 smooth items (daily staples), 45 intermittent items (specialty SKUs, niche flavors), 20 erratic items (trending products, weather-driven demand), and 5 lumpy items (large-format or B2B-adjacent).
A single model can fit some histories poorly. A simple average including zero days is a valid estimate of mean period demand; its problem is not that zeros inherently inflate or dilute the true mean. Compare candidate models using held-out errors and the business's ordering constraints.
For perishables, the consequences are more acute: over-ordering translates directly to spoilage. An intermittent item where SBA estimates 0.49 units/day versus Croston's 0.53 units/day sounds marginal in isolation. Across a 7-day order cycle with a 5% daily spoilage rate, however, the excess quantity compounds with the decay rate into real shrinkage. Evaluate intermittent-demand forecasting and the loss assumption separately; neither establishes safe shelf life.
Demand classification can inform model evaluation. Back-test candidate methods and review seasonality, events and changes in demand; no threshold eliminates the need for judgment or guarantees lower inventory at the same service.
How LineNow uses SBA
The current calculation classifies daily demand using ADI and CV² of non-zero demand. It uses SBA for intermittent and lumpy histories with sufficient observations. Fewer than six demand days are marked insufficient data and use a simple window-average fallback. Smooth and erratic histories use the window mean and daily variance in this calculation.
The SBA smoothing parameter in the current implementation defaults to 0.15. Its uncertainty estimate is derived separately from forecast errors; SBA itself is a point forecast, not a service-level or safety-stock guarantee. Counts, manual rates and other configured inputs can also affect the final recommendation.
Use the demand-pattern classifier to examine a series and the purchasing workflow to review a resulting order. Compare forecast errors on data not used for fitting, alongside stockouts and excess stock.
Related
Coefficient of variation (CV²) — the second axis of the SBC classification; determines whether SBA or standard smoothing applies
PAR level — the target inventory level built on the SBA-derived demand estimate
Safety stock — the statistical buffer above the demand estimate, computed separately from the SBA forecast
Decay rate — spoilage adjustment layered on top of the SBA estimate for perishables
Reorder point — the inventory level that triggers a new order; uses the SBA-derived consumption rate
