I'll point up the main issues for you:
-counterparty risk (if they don't pay or can't pay you're pretty much screwed)
-lack of certainty in the negative balance protection - specifically you're vulnerable to claims of "set off" if you are discovered to be the beneficial owner of all the accounts
-you're likely on the wrong side of AML regulations operating multiple accounts in this way
-spread/transaction costs eating into your likely returns
I've been privy to something similar though - there was (is?) a way to do something similar in the UK with negligible credit/counterparty risk and support from the regulatory regime.
In the UK, we have a FTT on listed stock purchases. Bona fide market markers are exempt - it is only the general public, pension funds, etc who pay the 0.5% of the notional value of the stock. Some clever city boys found a way around this, whereby a market maker / dealer who was exempt from the tax would write an OTC contract with a customer for the difference in value between notional purchases and sales of the stock. The customer, who would have been liable to 0.5% of the sale proceeds in stamp tax paid nothing as the contract with the dealer didn't involve ownership interests (or voting rights) in the underlying security. This arrangement came to be known as "contracts for difference".
It has been used for various interesting schemes - including taking the floating supply of a stock off the market by proxy, avoiding reporting requirements for owning x% of the issue, and insider trading both before and after legislation was updated to allow prosecutions where the person transacting in CFDs knew, ought to have known, or was reckless to the likelihood that their counterparty would open a corresponding position in the underlying market. Some Europeans involved with the recently publicised insider trading scheme used these products.
A feature of CFDs is that only a percentage of the contract value is required by the dealer as security for the contract, similar to margins on futures. This requires, when dealing with retail customers, a risk disclosure that losses may exceed deposits. In order to attract more business from the general public, one of the largest firms offering CFDs to retail approached the regulator for consent to offer "guaranteed stop losses" where the firm would bear any slippage risk if the customer stop loss order was triggered. Effectively negative balance protection by another name. The regulator allowed this as a specific and regulated contract of insurance and imposed ring fencing and reporting requirements on firms offering this type of product.
This insurance is expensive, often costing 10-15x the DV01. However, when there is event risk such as in FX over weekends, or more relevantly company earnings reports, constructing a virtual flat position by taking opposite positions with two dealers and two sets of "guaranteed stops" could be profitable. You've essentially paid to own a call and a put and should value the transaction accordingly. Sometimes this synthetic option position is cheaper than it should be.
Key elements:
- legality - you are dealing in regulated products- both the CFD and insurance portions - so there is certainty of result legally speaking and no "grey area" about exploitation
-you have accounts in your own name with 2 or more of the largest/creditworthy dealers
- creditworthiness - you're dealing with e.g. IG Markets and CMC markets who are listed blue chip firms
- insurance - your deposits are government guaranteed up to £85,000 per account holder in the same way as bank accounts
-you are paying 2 lots of spread, plus two premiums for the insurance - the expected move in the underlying needs to be enough to cover this
-you are exposed to some market risk legging into the positions and it is possible that one side may be triggered before the market close
-you are at the mercy of the dealer to accept your request to transact both to initiate and liquidate, and they'll back off their quotes if you are doing any size, in fast markets, of if they notice a pattern that you seem to be exploiting them - but in the main they are happy to quote smaller size and are protected by the aggregate premiums they receive and the spread
This used to be a worthwhile strategy, but many firms won't offer this insurance on individual equities any more, and on futures/FX they've defensively priced it. There are still occasional opportunities (e.g. the CHF this year) but it isn't something I pay attention to. Since about the beginning of the decade these firms have insisted that you eat the first part of the risk yourself (e.g. at least 1 big figure on FX) so it is less useful than it used to be.
The exception would be a smaller firm offering better deals* in the hope of attracting clients - this would be a shoddy offer due credit risk if the state didn't guarantee deposits up to £85k to any firm with a financial services licence whether they're undercaptialised for their risks or not. So there might be opportunity in the future. But no good for more than £50k or so.
* - the cost of acquiring virtual calls/puts is far cheaper than it should be, or can be arbitraged against another dealer
-counterparty risk (if they don't pay or can't pay you're pretty much screwed)
-lack of certainty in the negative balance protection - specifically you're vulnerable to claims of "set off" if you are discovered to be the beneficial owner of all the accounts
-you're likely on the wrong side of AML regulations operating multiple accounts in this way
-spread/transaction costs eating into your likely returns
I've been privy to something similar though - there was (is?) a way to do something similar in the UK with negligible credit/counterparty risk and support from the regulatory regime.
In the UK, we have a FTT on listed stock purchases. Bona fide market markers are exempt - it is only the general public, pension funds, etc who pay the 0.5% of the notional value of the stock. Some clever city boys found a way around this, whereby a market maker / dealer who was exempt from the tax would write an OTC contract with a customer for the difference in value between notional purchases and sales of the stock. The customer, who would have been liable to 0.5% of the sale proceeds in stamp tax paid nothing as the contract with the dealer didn't involve ownership interests (or voting rights) in the underlying security. This arrangement came to be known as "contracts for difference".
It has been used for various interesting schemes - including taking the floating supply of a stock off the market by proxy, avoiding reporting requirements for owning x% of the issue, and insider trading both before and after legislation was updated to allow prosecutions where the person transacting in CFDs knew, ought to have known, or was reckless to the likelihood that their counterparty would open a corresponding position in the underlying market. Some Europeans involved with the recently publicised insider trading scheme used these products.
A feature of CFDs is that only a percentage of the contract value is required by the dealer as security for the contract, similar to margins on futures. This requires, when dealing with retail customers, a risk disclosure that losses may exceed deposits. In order to attract more business from the general public, one of the largest firms offering CFDs to retail approached the regulator for consent to offer "guaranteed stop losses" where the firm would bear any slippage risk if the customer stop loss order was triggered. Effectively negative balance protection by another name. The regulator allowed this as a specific and regulated contract of insurance and imposed ring fencing and reporting requirements on firms offering this type of product.
This insurance is expensive, often costing 10-15x the DV01. However, when there is event risk such as in FX over weekends, or more relevantly company earnings reports, constructing a virtual flat position by taking opposite positions with two dealers and two sets of "guaranteed stops" could be profitable. You've essentially paid to own a call and a put and should value the transaction accordingly. Sometimes this synthetic option position is cheaper than it should be.
Key elements:
- legality - you are dealing in regulated products- both the CFD and insurance portions - so there is certainty of result legally speaking and no "grey area" about exploitation
-you have accounts in your own name with 2 or more of the largest/creditworthy dealers
- creditworthiness - you're dealing with e.g. IG Markets and CMC markets who are listed blue chip firms
- insurance - your deposits are government guaranteed up to £85,000 per account holder in the same way as bank accounts
-you are paying 2 lots of spread, plus two premiums for the insurance - the expected move in the underlying needs to be enough to cover this
-you are exposed to some market risk legging into the positions and it is possible that one side may be triggered before the market close
-you are at the mercy of the dealer to accept your request to transact both to initiate and liquidate, and they'll back off their quotes if you are doing any size, in fast markets, of if they notice a pattern that you seem to be exploiting them - but in the main they are happy to quote smaller size and are protected by the aggregate premiums they receive and the spread
This used to be a worthwhile strategy, but many firms won't offer this insurance on individual equities any more, and on futures/FX they've defensively priced it. There are still occasional opportunities (e.g. the CHF this year) but it isn't something I pay attention to. Since about the beginning of the decade these firms have insisted that you eat the first part of the risk yourself (e.g. at least 1 big figure on FX) so it is less useful than it used to be.
The exception would be a smaller firm offering better deals* in the hope of attracting clients - this would be a shoddy offer due credit risk if the state didn't guarantee deposits up to £85k to any firm with a financial services licence whether they're undercaptialised for their risks or not. So there might be opportunity in the future. But no good for more than £50k or so.
* - the cost of acquiring virtual calls/puts is far cheaper than it should be, or can be arbitraged against another dealer
