What information about the financial condition of firms is conducive to efficient and stable operation of the financial system and of the economy more broadly? In this essay, we outline the contours of an ideal set of such information, identify existing gaps and propose a way forward to fill them. We argue that an ideal set should comprise two dimensions. As regards financial characteristics, it should cover three different types, viz: estimates of the current financial condition ("first-moment information"); estimates of risk profiles ("risk information"); and measures of the uncertainty surrounding both kinds of estimate ("measurement error information"). As regards the object of the analysis, it should cover information about both the individual firm ("micro information") and, suitably aggregated, the "system" as a whole ("macro information"). So far, efforts have mainly focused on micro information and, within it, on estimates of the current financial condition; by contrast, risk information has drawn attention only more recently and measurement error information has been largely neglected. We also note that, as regards micro information, significant differences in perspective between accounting standard setters and prudential supervisors have come to light. We examine the reasons for these differences and propose ways in which they could be reconciled. We propose a strategy based on two principles: first, in the long term, the "decoupling" of the objective of accurate financial reporting about the firm from that of instilling the desired degree of prudence in its behaviour; and second, a "parallel" process towards that objective so that at all points the prudential authorities can neutralise any undesirable implications for financial stability of changes in financial reporting standards. We stress that close cooperation between accounting standard setters and supervisory authorities is called for both in developing the final set of information and in implementing it.