ARIMA need exogenous array

When I tried to follow the video and ran the following code, I have the corresponding error message and don't know how to fix it.   model_auto = auto_arima(df.ret_ftse[1:], exogenous = df[["ret_spx","ret_dax","ret_nikkei"]][1:],
m = 5, max_p = 5, max_q = 5, max_P=5, max_Q=5) df_auto_pred = pd.DataFrame(model_auto.predict(n_periods = len(df_test[start_date:end_date])),
exogenous = df_test[["ret_spx","ret_dax","ret_nikkei"]][start_date:end_date],
index = df_test[start_date:end_date].index)
df_auto_pred.plot(figsize =(20,5), color = "red")
df_test.ret_ftse[start_date:end_date].plot(color = "blue")
plt.title("Auto Model Predictions vs Real Data", size = 24)
plt.show()          

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-28-3266f5b3ac5e> in <module>
      3                        m = 5, max_p = 5, max_q = 5, max_P=5, max_Q=5)
      4 
----> 5 df_auto_pred = pd.DataFrame(model_auto.predict(n_periods = len(df_test[start_date:end_date])),
      6                             exogenous = df_test[["ret_spx","ret_dax","ret_nikkei"]][start_date:end_date],
      7                            index = df_test[start_date:end_date].index)

~\anaconda3\lib\site-packages\pmdarima\arima\arima.py in predict(self, n_periods, exogenous, return_conf_int, alpha)
    626 
    627         # if we fit with exog, make sure one was passed:
--> 628         exogenous = self._check_exog(exogenous)  # type: np.ndarray
    629         if exogenous is not None and exogenous.shape[0] != n_periods:
    630             raise ValueError('Exogenous array dims (n_rows) != n_periods')

~\anaconda3\lib\site-packages\pmdarima\arima\arima.py in _check_exog(self, exogenous)
    493         if self.fit_with_exog_:
    494             if exogenous is None:
--> 495                 raise ValueError('When an ARIMA is fit with an exogenous '
    496                                  'array, it must also be provided one for '
    497                                  'predicting or updating observations.')

ValueError: When an ARIMA is fit with an exogenous array, it must also be provided one for predicting or updating observations.