Resumen
This paper argues that a simple white noise process with one jump in its unconditional variance may give rise to the presence of ARCH effects, and, surprisingly, this may occur in determinate circumstances even when the jump is very brief. Though ARCH effects are not denied, this evidence, together with some empirical results obtained from Standard & Poor?s 500 returns, allows one to question whether they are a general and regular property of so many economic and financial series.Keywords: ARCH, autoregressive conditional heteroscedasticity, stock returns, unconditional varianceJEL Classification: G12