I think the big risk would be that you can't pay off the balloon payment in December due to a down market, but having a sizable emergency fund (which many people do) mitigates this.
This is exactly what you SHOULD NOT DO. Don't invest your payments. Our strategy is basically, put all expenses on to the cards, maxing out our accounts to the point where contributions go in equally until October. This leaves us with enough cash each month to pay the mortgage and two cash required utility bills and enough to dump into savings so that 2017 IRA's are ready by October. October-December we will stash cash to make the payments in January on the first two cards and hold and save the cash for the Slate payment in May. When the 0% expires the cards are paid off and will be closed, then DW and I will switch which of us opens the new cards.
We do this to front load multiple (457/403)*2 retirement accounts. Maxed by October 1 using 0% Amex Blue Cash Preferred and 0% Citi Double Cash Back and moving the balance to Chase Slate. The first two cards have 0% til February 2017 and Chase until May. Chase Slate is 0% balance transfer fee for any transfers for the first 60 days. Combining cash back and interest (for 2017 IRAs) in 1% savings, nets about 3.5% based on our spending*, not including returns/dividends on the shares purchased.
Who needs coupons when you have math? At a minimum this wipes the interest on our mortgage.
Our mortgage is 2.875% with 14 years remaining, and yearly isn't more than the standard deduction. We should be FI in 5 years and planning to FIRE in 8. Contribution/Savings Rate does more for our goal then expected growth. We cannot control the return rate, but we can control to some extent how much is there to grow. Our jobs are extremely stable, so our tolerance favors putting math to work for us as long as possible. The flat market/volatility for the past 15 months has prompted us to contribute as much as possible early in our FIRE journey, we started last January.
A couple more thoughts...
It's not for the risk averse, people with unstable income, or people who SHOULD have an emergency fund.
It's not sustainable. You have to weigh the risks of carrying over a balance to a new card (and possibly a new year) if you can't pay it off. You have to know for yourself how much you're comfortable rolling over year after year or when you might have to slow the retirement contributions. You're borrowing against time and your ability to pay it off quickly when the time comes. Getting in the market sooner for longer beats not investing for a short time to pay off cards in the future. I'm comfortable in a scenario where we do this for a couple years, get half way through a year of front loading and then turn off contributions for a month or two, pay off the current card, and continue contributions as long as we can still max out the deferred savings by year end (for tax savings and buckets you don't get back). The goal is to still pay them off completely by end of the 0% term, but I could see getting comfortable rolling balances as long as we can get 0% monthly and on balance transfers.
*Amex Preferred Blue Cash 6% Groceries, 3% Gas, 1% everything else, except we use Citi to get 2% cash back. Discover 5% gas for 3 months. Most of our card spending is groceries and gas, and we can churn the 6% for gift cards for some other purchases.