Diagram 4 describes the cumulative net cash flow in the same way as in Diagram 3, but aggregated here for the five large banks in Sweden. The diagram data is based on a monthly average for the period March 2018 to December 2021 inclusive.

The graph reaches its local maximum exactly for the time bucket that includes day 30, which means that the banks’ cumulative net cash flows will be most positive after 30 days. This supports the theory that the banks focus on attaining a lower liquidity risk on day 30 to optimise LCR, while the liquidity risks, at least according to the reported contractual net cash flows, are higher on the days before and after that day. In the same way, the graph reaches its local minimum between day 30 and one year. This supports the theory that the banks also focus on achieving a low liquidity risk at one year when NSFR is measured, while liquidity risks are higher in the interim period. Taking a monthly average for the periods evens out more extreme values. It thus hides the fact that there have been times when the local minimum and maximum values for one or several banks have been much more obvious, which could imply higher risk (see for example in Diagram 6 in the Appendix how the average for different time periods can shift over time).^[15] The mean for the local maximum point is greater than the mean for the local minimum point in Diagram 4 with more than a 99.9 percent probability (t-value = 6.41).

Figure 4. Cumulative contractual net cash flows, aggregated for the five large banks in Sweden, monthly average (SEK billion)

All banks are obliged to fulfil LCR and NSFR, which reduces liquidity risks in the banks. It is possible that the banks fulfil these requirements by reducing liquidity risks at the very points in time of around 30 and 360 days, while liquidity risks in other time buckets are as high as before LCR and NSFR were introduced. It is also possible that the banks redirect their cash flows such that liquidity risks have actually increased in other time buckets. In the absence of relevant data for the period before LCR was introduced, this question cannot be answered here. In any case, there is a risk that the banks have liquidity risks that these metrics do not capture. A further risk is that many banks apply similar liquidity risk optimisation, and will thus also be most vulnerable at the same points in time. It is therefore important to monitor net cash flows for all time periods and not just for 30 days and one year. This is the case not only for the banks themselves, but also for those tasked with assessing and identifying vulnerabilities in the financial system. Liquidity risks can also arise in different currencies, and it is therefore important to also calculate DLC for significant currencies.