Now, before you skip this post as crazy, hear me out. The question here is whether banks may look at the capital charges for equities securities loans done as principal for their hedge fund clients as too expensive under Basel III. This may encourage banks to want to get out of the principal game and into an agency model.
Read below for my thesis on why “agency securities lending” might be considered realistic.
I invite you to follow my train of thought to get to a model of agency securities lending:
-> To start with, Basel III’s new ratios, particularly the Liquidity Coverage Ratio, mandate that a bank must have a certain amount of High Quality Liquid Assets relative to expected cash outflows over the next 30 days.
-> In an equity securities loan that turns over daily (no term loans in this thesis), there are relatively stable amount of assets to expected outflows.
-> BUT, equities are not considered High Quality Liquid Assets (HQLA) under Basel III. This means that the denominator gets larger as a bank lends out more as principal, but the numerator will not increase in size because the equities aren’t considered high quality.
-> This scenario would cause a bank’s Liquidity Coverage Ratio to be outside an accepted range, and this is not going to happen as far as any risk manager is concerned.
In my thesis, instead of acting as principal and pushing their Liquidity Coverage Ratios, banks would act to identify loans at agent lenders on their hedge fund clients’ behalf and sponsor clients to make their borrows. Banks do a version of this now with the CCPs out there for securities lending, but my hypothesis takes the conversation in a different direction. I am thinking that a bank will arrange for its largest hedge funds that they will be approved as principal borrowers by agent lenders themselves.
Why would the agents agree to this? Because the banks would offer indemnification to the agents against losses in the borrow book. Their capital charges for indemnification will be much lower than the charges for not having HQLA meet their net cash outflows. An indemnification charge would affect the Leverage Ratio, not the Liquidity Coverage Ratio. In fact, the way the current conversation is going, they may not have any charge for their indemnification.
Now, I reserve the right to make all these points moot and retract this post the minute that a smart accounting type shows me the error of my ways, FASB 39 gets rewritten or Basel IV comes out, but in the meanwhile, I am putting this idea out there for consideration.