SKEW Index
Calculate a SKEW index with the Cboe SKEW Index methodology, for any ticker with listed options: the risk-neutral skewness S of the log return over the coming days, read from the prices of out-of-the-money options and expressed as 100 - 10 S. A SKEW of 100 means the options price a symmetric distribution; the higher the index, the more the options price a large fall (tail risk). The Cboe SKEW of the S&P 500 has mostly ranged between 110 and 150.
The formula is as follows, per expiry with \(R = \ln (S_{T} / F_{0})\), \(y_{i} = \ln (K_{i} / F_{0})\) and the strikes \(K_{i}\), prices \(Q (K_{i})\) and intervals \(\Delta K_{i}\) of the VIX methodology:
\[P_{1} = - \frac{e^{rT} \sum Q(K_{i}) \Delta K_{i}}{K_{i}^{2}} + \epsilon_{1}\] \[P_{2} = \frac{e^{rT} \sum 2 (1 - y_{i}) Q(K_{i}) \Delta K_{i}}{K_{i}^{2}} + \epsilon_{2}\] \[P_{3} = \frac{e^{rT} \sum 3 (2 y_{i} - y_{i}^{2}) Q(K_{i}) \Delta K_{i}}{K_{i}^{2}} + \epsilon_{3}\] \[\epsilon_{1} = - \left(1 + \ln\left(\frac{F_{0}}{K_{0}}\right) - \frac{F_{0}}{K_{0}}\right)\] \[\epsilon_{2} = 2 \ln\left(\frac{K_{0}}{F_{0}}\right) \left(\frac{F_{0}}{K_{0}} - 1\right) + \ln\left(\frac{K_{0}}{F_{0}}\right)^{2}\] \[\epsilon_{3} = 3 \ln\left(\frac{K_{0}}{F_{0}}\right)^{2}\left(\frac{\ln\left(\frac{K_{0}}{F_{0}}\right)}{3} - 1 + \frac{F_{0}}{K_{0}}\right)\] \[S = \frac{P_{3} - 3 P_{1} P_{2} + 2 P_{1}^{3}}{(P_{2} - P_{1}^{2})^{\frac{3}{2}}}\]and over the near and next term around days:
\(P_{1}\), \(P_{2}\) and \(P_{3}\) are the prices of R, \(R^{2}\) and \(R^{3}\), the \(\epsilon\) terms correct for the expansion around the forward price \(F_{0}\) while the puts and calls are split at \(K_{0}\), the first strike at or below it, and r is the continuously compounded risk free rate. The terms, the forward price, \(K_{0}\) and the strikes used are those of the VIX methodology (see get_model_free_implied_volatility), as the white paper prescribes. \(\epsilon_{2}\) holds the full \(\ln (K_{0} / F_{0})^{2}\) of the expansion of \(R^{2}\) around \(K_{0}\), where the white paper prints half of it (see market_metrics_model.get_cboe_skewness); for index options the difference is negligible. Unlike the variance of the VIX, the skewness is interpolated linearly in time. Applied to the S&P 500 index options (“^SPX”) with days=30 it approximates the Cboe SKEW Index, which uses the Cboe’s own quotes and Treasury bill rates per expiry.
For more information, see the following papers:
- Cboe Global Markets (2010). “The Cboe SKEW Index - SKEW.” White paper.
- Bakshi, G., Kapadia, N. and Madan, D. (2003). “Stock Return Characteristics, Skew Laws, and the Differential Pricing of Individual Equity Options.” The Review of Financial Studies, 16(1), 101-143.
Also known as: Cboe SKEW, tail risk index, black swan index, implied skewness index.
SKEW Index in Python or via the MCP
Install the Finance Toolkit with:
pip install financetoolkit -U
Then call get_skew_index as shown below.
from financetoolkit import Toolkit
toolkit = Toolkit(["^SPX", "AAPL"], api_key="FINANCIAL_MODELING_PREP_KEY")
toolkit.options.get_skew_index(
show_columns=["Near-Term Expiration", "Next-Term Expiration", "Skewness", "SKEW Index"]
)
Or, after setting up the Finance Toolkit MCP, ask:
What is the SKEW index of the S&P 500 and Apple, and which near-term and next-term expirations is it based on?
Which returns:
| Item | Near-Term Expiration | Next-Term Expiration | Skewness | SKEW Index |
|---|---|---|---|---|
| ^SPX | 2026-12-18 | 2026-12-21 | -4.2626 | 142.626 |
| AAPL | 2026-12-18 | 2026-12-24 | -1.7288 | 117.288 |
Parameters
get_skew_index accepts the following parameters.
| Parameter | Type | Description | Example |
|---|---|---|---|
days | int | The constant maturity in calendar days. Defaults to 30. | 360 |
rounding | int | None | The number of decimals to round the results to. Defaults to None, which uses the Toolkit’s rounding. | 2 |
show_columns | str | list[str] | None | The columns to keep, a name or a list of names, matched on every column level, or the rows when no column has those names (such as the line items of a financial statement). Names that do not exist are logged, and when none exist the result is returned in full. Defaults to None, which keeps every column. | None, which keeps every column |