Would This Repo+CDS Strategy Work?

So, up front I admit that I have limited knowledge of the CDS market and even less of the repo market. But I thought of a strategy today that could hypothetically work with certain assumptions. The crux of the strategy is to attempt to locate inefficiencies in the CDS and Repo markets.

The strategy would be to purchase a high grade corporate bond, purchase a CDS and enter into a repo to replicate the transaction. The idea would be for the interest from the coupon to be in excess of the costs of the Repo and CDS and then replicate the position multiple times to get an attractive return with little risk.

For example, say we bought $100MM of a 5% A rated bond at par, bought a CDS on it for, say, 100 bps (annualized) and then entered into a repo on very conservative terms of 3.0% with a 20% haircut. We're generating 5mm from the coupon and paying 1mm for the cds and 2.4mm from the repo (0.03*80MM), leaving us with 1.6MM in cashflow. We then replicate the transaction with the 80MM we received from the repo.

So, I used a lot of assumptions here, most notably pricing. It seems logical to me that you could enter a repo agreement at a rate lower than the coupon on a bond due to credit enhancement, by providing collateral at 125% of the repo amount and the cds, it seems likely that the spread between the repo and a comparable treasury would be very tight. Adding the cds, depending on the cp, could by itself upgrade the bond in theory (similar to muni bond ratings vs. their underlying ratings). It's a play on the combination of inefficiencies through CDS and repo pricing and by levering up via the repo you effectively lock in a large return with virtually no risk, excluding the possibility of a double default of course. The strategy seems like it would work best in an environment with larger spreads, allowing the potential for more inefficiencies.

What I am unsure of is the term of repos. I know most are overnight, but aren't there term repo's as well for longer horizons (e.g. 1 or 2 years)? If not you would have interest rate risk on the bonds and repricing risk on the repo's. Also I am assuming the repo market is somewhat inefficient because of the uniqueness of them. If I am not mistaking they aren't standardized, so couldn't having the combination of an appropriate haircut, repo rate and cds pricing create enough inefficiencies to provide arbitrage opportunities. Also, I know the CDS market has became more efficient recently, but I assume there are still some opportunities.

I'm sure I am wrong in these assumptions, because if not you could create huge returns with virtually no risk and the market would recognize so quick that the opportunity would diminish quickly. I'm just curious where the flaw in this strategy is. If nothing else, I would think that creating a strategy using as many inefficient instruments as possible could create opportunities, right? Anyways, I'm a value guy and never really get interested in any derivative outside of equity options, but I thought of this today and was curious what was wrong with my thinking on it.