ok so a little wonky but prepayment speeds on 2020-2021 low-coupon MBS are *extremely* low. partly because they are seasonal and the last few months are the seasonal lows. these securities are priced at a discount to par because interest rates are way up. these MBS are the largest issuance in the market because so many people refinanced at low rates. that means that the securities that banks hold are backed by these loans, but they don’t necessarily own the entire mortgages.

i get the decision by tesla to do price cuts to gain share but thinking like an underwriter (sorry, the brainworms are terminal) I would want lower LTVs if, on top of normal depreciation and new -> used step down, you need to account for falling prices on new models. of course that’s not really the case for buyers of new teslas, who are overwhelmingly prime borrowers, but it would lower recoveries all else equal. i suspect that will make it a hard make to lend against in the used + non-prime market.

Since I keep getting tags and DMs about the shortfall in tax payments not-withheld I will give you the best thing I have so far. the blue line is the daily nowcast for the total of cumulative tax collections between june 1st of prior year and april 30th of current year the idea is basically that we use estimated tax payments to... wait for it... estimate tax payments. groundbreaking, i know, but it's useful. The shortfall between the may 31st nowcast and whatever the april 30th total ends up being is probably mostly the California effect

to clarify a question from the DMs:
high realized vol in FI means an equal-weighted portfolio has less bonds, negative correlation means more leverage to reach vol targets or before reaching vol caps. the opposite of pre-SVB.

https://twitter.com/NewRiverInvest/status/1648074791260413952

dispersion from outside effects (macro, data, sentiment, fundamentals) -> deviation from targets -> rebalancing -> mean reversion -> reinforcement of (neg) correlation. its like this time last year, but in reverse. the pressures are to re-gross.

so it's been almost 3 weeks now since the timer crowd started talking about how "PUTs are always cheapest when you need them most" and since then there's vega losses on top of the decay. the anchoring to Dec/Feb is going to keep people burning premium expecting a continuation of the previous range & revisiting of highs & lows bc they are not recognizing the big change: stock-bond correl went negative. this environment is not comparable to the pre-SVB environment.