Why futures implied repo spreads may not be arbitrage

# Is my thinking on futures implied repo correct?


# Is my thinking on futures implied repo correct?












I am building analytics for futures and have a theoretical understanding. If implied repo > actual repo then I can short futures and go long the security and finance it in repo.

ON my Bloomberg screens for the 2-yr contract (March 2020), I see the implied repo is about 25 basis point above the actual repo. Isn't this a no brainer then? I am guaranteed to earn 25 basis points if I hold the position to expiry.

Wouldn't market participants jump on this such that this opportunity goes away? The other way to look at it is that the net basis is also negative (-2 to -3 ticks). Assuming the CTD doesnt switch, it should converge to zero so easy money?

I must be missing something.

## Answer by dm63 (score 4)

https://quant.stackexchange.com/a/49606

By coincidence I was looking at this yesterday. The implied repo on the TUH0 contract is about Fed funds + 38, which indeed is around 25bp higher than the term repo to March. Why does this apparent arbitrage exist ? The answer is that consumes financial resources- in particular it will be an on-balance sheet transaction. There is a risk that these resources will be taken away from you at some point during the trade. In particular, the trade goes over year end, when banks try to reduce their balance sheet. The effect of this is to reduce the number of banks able to participate in the trade, and it also reduces the number of non banks because these need financing from the banks. Why do banks reduce balance sheet over year end? Basically to make the bank look smaller for a number of regulatory and accounting reasons.

Bottom line : go ahead with the trade, but ask your boss whether she can guarantee the balance sheet will not be taken away for year end.

## Answer by Attack68 (score 1)

https://quant.stackexchange.com/a/49617

#### Capital Requirements

In addition to @dm63 answer, I recall the primary concern in the repo space was the impact to the 'exposure measure' in the regulatory Leverage Ratio (LR). Globally Systemic banks (in different jurisdictions) have historically had relatively tough minimum requirements for the LR (6% in the US), which leaves a 2 or 3 tick profit (or 200k per 1bn notional), regardless of the maturity of the bond, particularly unappealing, especially compared to other potential allocations of the capital.

If you consider the return on capital employed, since 6% of capital is required per exposure (in the worst case) this trade (if the exposure is increased by 1bn) requires 60mm capital. The return of 200k is 0.33%.

Even if you take the best case where only 3% of capital is required and the return is 0.66%. This is still pretty useless: I have seen an investment bank turn down trades which yield 5% return on capital regularly.

#### Technical Requirements

From what I remember about net basis trading (going back 10 years) it was also something that many people struggled to fully appreciate the nuances of managing the risk. If I remember correctly you had to calculate the specific amount of futures that were needed on an ongoing basis (i.e. to hedge MTM) and then going into settlement (EDSP) you needed to trade additional futures lots to ensure your notionals for delivery matched.

If you get that wrong those 2 or 3 ticks disappear away fairly easily.