According to a survey run among primary dealers by the US Treasury Department in January 2018, the US fiscal deficit would account for $750bn in 2018, $965bn in 2019 and $1025bn in 2020 (far above the numbers indicated in the latest CBO projections made in June 2017, ie before the adoption of the fiscal reform). Given the size of the future Fed’s non-reinvestments ($229 bn in 2018, $270 bn in 2019 and $182 bn in 2020 according to our calculations), the net marketable borrowing would account for $979 bn in 2018, $1235 bn in 2019 and $1207 bn in 2020. Considering that 30% of the net issuance of govies has corresponded to T-bills over the past 12 months and making the assumption that this share will remain the same in the future, the scenario of primary dealers would imply that the net issuance of long-term US Treasuries would account for $685 bn in 2018, $864 bn in 2019 and $845bn in 2020. This represents a rather big increase compared with 2017 (the net issuance of long-term US Treasuries had been $399 bn) but remains below the amounts of long-term US Treasuries that the non-Fed investors had to absorb over the past few years (for instance, net purchases of non-Fed investors accounted for $1.6 trn in 2010 and $1 trn in 2012).
Less than the absorption of the net issuance of long-term govies in 2018 and 2019, the worrying thing for investors is that this an anomaly to see deficits climbing so much at this point of the cycle (after a very long expansion and with a very low unemployment rate) as there has not been a rise of deficits coupled by a decrease in the unemployment rate during the past decades.