Hausman-Wu Test

Calculate a regression-based Hausman-Wu test for the endogeneity of suspect_ticker in a regression of dependent_ticker on suspect_ticker (and, optionally, other_independent_tickers), using instrument_tickers as instruments for suspect_ticker.

OLS requires every regressor to be uncorrelated with the error term. When that fails, for example because of an omitted common driver, reverse causality or measurement error, the regressor is “endogenous” and OLS is biased; an instrumental-variables approach such as get_iv_2sls is then needed. This test checks whether that is the case before committing to a full IV fit, using two auxiliary OLS regressions:

1. Regress suspect_ticker on the instruments (and the other tickers, if given) and keep the residuals: the part of suspect_ticker that the instruments do not explain.
2. Regress dependent_ticker on suspect_ticker, those residuals (and the other tickers) and test whether the coefficient on the residuals is zero.

If suspect_ticker is exogenous the residuals add no information, so their coefficient should be statistically zero. A low p-value is evidence that suspect_ticker is endogenous and that the OLS estimate is unreliable. The test is only as good as the instruments: they must be correlated with suspect_ticker but not with the error term.

The formula is as follows:

• All variables are the tickers’ column (Return by default)

For more information about the method, see the following papers:

• Wu, D-M. (1973). “Alternative Tests of Independence Between Stochastic Regressors and Disturbances.” Econometrica, 41(4), 733-750.
• Hausman, J.A. (1978). “Specification Tests in Econometrics.” Econometrica, 46(6), 1251-1271.

Also known as: Hausman test, Durbin-Wu-Hausman test, regression test for endogeneity.

Hausman-Wu Test in Python or via the MCP

Install the Finance Toolkit with:

pip install financetoolkit -U

Then call get_hausman_wu_test as shown below.

from financetoolkit import Toolkit

toolkit = Toolkit(
    ["SMCI", "NVDA", "TSM"],
    api_key="FINANCIAL_MODELING_PREP_KEY",
    start_date="2025-01-01",
    end_date="2025-12-31",
)

toolkit.econometrics.get_hausman_wu_test(
    dependent_ticker="SMCI", suspect_ticker="NVDA", instrument_tickers="TSM"
)

Or, after setting up the Finance Toolkit MCP, ask:

Which returns:

Using Your Own Data

get_hausman_wu_test runs on any data, not only prices: ratios, economic indicators, yield curves or your own dataset. Pass a DataFrame, a Series or a dictionary of them to Econometrics as data, with the dates in the rows or the columns, and set changes to how its changes are calculated: "percentage", "log", "difference" for rates and ratios, or None when the data are changes already. For example, for the monthly unemployment rates of the United States and Germany, where each country is a column:

from financetoolkit import Economics, Econometrics

economics = Economics(start_date="2000-01-01")

unemployment = economics.get_unemployment_rate(
    countries=["United States", "Germany"], period="monthly"
)

econometrics = Econometrics(data=unemployment, changes="difference")

Then call econometrics.get_hausman_wu_test with dependent_ticker, suspect_ticker and instrument_tickers, using the column names of the data as tickers.

Parameters

get_hausman_wu_test accepts the following parameters.

Related Specification Tests