Some thoughts on Ando's theorem and Parrott's example (Q5957207)

\textit{T. Ando}, in his paper, `On a pair of commutative contractions' [Acta Sci. Math. 24, 88-90 (1963; Zbl 0116.32403)], proved if \(T_{1}\) and \(T_{2}\) are two commuting contractions on a Hilbert space \(H\), then there are two commuting unitary operators \(U_{1}\) and \(U_{2}\) on a Hilbert space \(H_{0} \supset H\) such that \( T_{1}^{m} T_{2}^{n} (x)= P_{H} U_{1}^{m} U_{2}^{n}(x) \), for every \(m\)=1, 2, \dots , \(n\)=1, 2, \dots and \(x \in H\), \(P_{H}\) being the projection of \(H_{0}\) on \(H\); this is called Ando's dilation theorem. A well-known consequence of this theorem is every contractive homomorphism on the bi-disc algebra is automatically completely contractive. \textit{S. K. Parrot}, in his paper `Unitary dilation of commuting contractions' [Pac. J. Math. 34, 481-490 (1970; Zbl 0202.41802)], gave an example to prove that this result cannot be extended to the tri-disc algebra. In the present paper, by analyzing Parrott's example, the authors give an example of a contractive homomorphism on the tri-disc algebra which is not even 2-contractive.