U.S. All Commodities Producer Price Index

USPPI is the producer price index that measures the average change over time in the selling prices producers receive for a fixed all-commodities basket at the first commercial transaction. The recorded raw level (1982 = 100) is carried through without any transformation, and because no seasonal adjustment is applied it is the unadjusted all-commodities headline level rather than a core variant.

As an aggregate price index, its level follows the base-period-quantity-weighted fixed-basket construction (Laspeyres 1871), which is paired with the comparison-period-weighted form (Paasche 1874). The two forms are reconciled by the geometric ideal index and the expenditure-share-weighted log-change form (Fisher 1922; Törnqvist 1936).

The economic-theoretic object this index approximates is the expenditure or revenue ratio with the underlying program held fixed. It sits within the cost-of-living tradition and the output-price reading systematized by the four approaches (Konüs 1939; Diewert 2002). Whether a given formula is exact or superlative for an aggregator function is settled by the ideal-index literature (Diewert 1976). The axiomatic consistency a measured index satisfies is surveyed in the literature, which also sets out the substitution wedge between a fixed-base index and its theoretical target (Balk 1995; Diewert 1998).

The persistence the recorded number shows around its trend is interpreted within the inflation-measurement frame (Cecchetti 1997), and the upward fixed-base measurement bias this index inherits is catalogued in the literature (Boskin et al. 1996). Since the series is unadjusted, its within-year pattern is left intact, so USPPI is read as the unrevised level whose growth rate is producer-price inflation (Stock and Watson 2007).

KRED applies no transformation to USPPI and stores the index exactly as compiled, performing no rescaling, deflating, smoothing, annualization, or seasonal adjustment. The methodology is therefore that of the underlying index-number construction rather than any KRED computation.

The level is built as a base-period-weighted aggregate of price relatives anchored at 1982 = 100 (Laspeyres 1871), which is paired with the comparison-period-weighted form (Paasche 1874). Reconciling the two through the time- and factor-reversal tests yields the ideal index (Fisher 1922). The superlative index that is exact for a flexible aggregator, namely the log-change index and the family it belongs to, supplies the benchmark against which the fixed-basket formula is judged (Törnqvist 1936; Diewert 1976).

Which reversal and consistency properties a formula must hold is fixed by the axiomatic or test approach, which sits within the four-approach taxonomy (Balk 1995; Diewert 2002). The substitution and quality gap between the compiled fixed-base aggregate and the economic-theoretic index is quantified against the cost-of-living target (Konüs 1939; Diewert 1998; Boskin et al. 1996).

No seasonal filter is applied, so the recorded level carries its within-year price pattern intact and the decomposition into trend, transitory disturbance, and bias is left to the user (Cecchetti 1997). For an analyst who summarizes the cross-sectional dispersion of the component price changes, the limited-influence estimators and the kurtosis-optimal trimmed mean describe how an underlying-inflation reading is obtained from the same recorded basket, and their derivation implies a heavier per-tail trim for producer prices than for consumer prices (Bryan and Cecchetti 1994; Bryan, Cecchetti, and Wiggins 1997).

As a producer-price gauge, USPPI carries the upstream inflation signal for the cost pressure facing firms before it reaches the consumer stage. This distinction between upstream and downstream prices rests on the cost-of-living target and the index-number scaffold (Konüs 1939; Diewert 2002).

Because the recorded level embeds transitory supply shocks alongside the persistent trend, its interpretation rests on the trend-plus-noise-plus-bias decomposition and the unobserved-components trend extraction (Cecchetti 1997; Stock and Watson 2007). Which part of the move is fundamental is framed by the long-run output-neutral notion of core inflation and the original supply-shock-free concept (Quah and Vahey 1995; Eckstein 1981).

The asymmetry of the cross-sectional price-change distribution makes the headline producer-price move sensitive to a few large relative-price shocks. Because it is handled by the limited-influence estimators, the analyst separates broad pipeline pressure from sectoral noise (Bryan and Cecchetti 1994; Bryan, Cecchetti, and Wiggins 1997).

The fixed-basket substitution and quality biases caution that the measured producer-price level overstates true price change (Boskin et al. 1996; Diewert 1998). The axiomatic reading and the base-weighted construction explain why the gauge moves mechanically with the chosen weights when input shares shift (Balk 1995; Laspeyres 1871), and because the series is unadjusted its month-on-month change has to be read together with the within-year pattern.

For market participants, USPPI is the upstream margin-and-inflation input that feeds breakeven and discount-rate views. Because producer-price inflation often leads the consumer measure, it is read through the underlying-inflation lens and the trend extraction (Cecchetti 1997; Stock and Watson 2007).

Pricing real and nominal claims requires separating the persistent producer-price trend from the transitory component, which the core-inflation estimators and the kurtosis-optimal trim provide (Bryan and Cecchetti 1994; Bryan, Cecchetti, and Wiggins 1997). The heavier producer-price trim reflects the fat tails of the input-price distribution, and the supply-shock-free concept and the output-neutral core define the signal a fixed-income desk discounts (Eckstein 1981; Quah and Vahey 1995).

Because the quoted level is a base-weighted fixed-basket aggregate reconciled by the ideal index and the log-change form (Laspeyres 1871; Paasche 1874; Fisher 1922; Törnqvist 1936), its mechanical sensitivity to weight choice and the substitution and quality biases matter for any contract or swap indexed to producer prices (Boskin et al. 1996; Diewert 1998).

The axiomatic constraints and the four-approach foundation tell a trader which index properties are preserved when the series is used to deflate revenues or to set escalation clauses (Balk 1995; Diewert 2002), and since the level is not seasonally adjusted, settlement comparisons have to be aligned on the same calendar month.

USPPI is published as a not-seasonally-adjusted all-commodities producer-price index level, and the appropriate statistical object is its logarithm, expected to be integrated of order one. The sample is 365 monthly observations spanning 1996-01 to 2026-05, and the unit-root triplet is fitted on the log level with a constant and trend.

The augmented Dickey-Fuller regression of Dickey and Fuller (1979), in the lag-augmented form of Said and Dickey (1984), does not reject the unit root at p = 0.295, the nonparametric test of Phillips and Perron (1988) agrees at p = 0.431, and the stationarity-null test rejects trend stationarity at the five percent level with p = 0.011 (Kwiatkowski et al. 1992), so all three procedures concur on a clean I(1) classification.

On a clean I(1) reading the level is not the stationary object, so the portmanteau and the break search are run on the first difference of the log level, since a level portmanteau reads the near-unit autocorrelation of a stochastic trend and a level mean-break search spuriously segments it (Perron 1989; Hamilton 2018). The portmanteau statistic of Ljung and Box (1978), refining the form of Box and Pierce (1970), rejects the white-noise null on the differenced log level at p = 0.000, while the multiple-break procedure of Bai and Perron (1998), computed by the dynamic-programming algorithm of Bai and Perron (2003), finds no break in the mean of the differenced series.

Seasonal tests are appropriate on this unadjusted monthly index. The seasonal-unit-root test of Hylleberg et al. (1990), with the monthly mechanics of Beaulieu and Miron (1993), rejects unit roots at the seasonal frequencies at p = 0.000, while the seasonal dummies are jointly significant at p = 0.000, so the seasonality is present and deterministic rather than stochastic, consistent with the stationary-seasonality reading of Canova and Hansen (1995) and with the not-seasonally-adjusted character of the series (Ghysels and Osborn 2001).