macrocephalopod

What method tends to be the easiest to back out ivols from American, dividend-producing options? I've been using broker ivols for a while now, but I'd like to get my hands dirty building an SSVI surface directly from option prices.

Step 1 was going to be using a good risk-free rate for the option lifetime. The general consensus seems to be an IBOR, so FRED has that and I can match the option tenor to the nearest rate possible there. It's kind of gone now, so I was going to use T yields too/ interpolate.

Step 2 - the hardest part seems to be the dividend yields. You can of course for like 95% of cases take the discrete dividends and make a continuous implied yield, but you need a dividend forecast. In practice is it fine to use the historical trailing dividend yield (12mo)?

I guess this is a question mostly directed at folks like @TheSpeculator0 @bennpeifert @darjohn25 Another idea would be to use comparables (e.g. dividend changes over the company peer group) or pay for a fancy dividend forecasting service, but at that point I just give up and use

broker ivols anyway.

I know you can hack using the forward -prices to skip the interest rate and dividend yield nonsense since that's baked into the price of the forward itself, but there isn't (at least any exchange-traded) forwards on single stocks, just the indices.

always compute forwards first from put/call parity bounds

(don't mess with dividend forecasting etc, like, the market tells you the answer, and if you're different from the market, you're almost certainly wrong)