Subject: Re: OT: Car Max (KMX)
Is it better to roll on an up day? A down day?

Ideally you want to
* Roll as seldom as possible, since the bid/ask gaps are horrible making trading expensive.
So, generally it's better to roll every two years than every one year.

* Within the constraint above, roll at an early date.
Time value evaporates more and more rapidly as it gets closer to expiry.
So avoid the last 3 months if you can.
In particular, if you sell less than around a month before expiry the bid/ask gap is often bigger than the remaining time value,
so you end up not even being able to sell for the in-the-money value. This is definitely something to be avoided.
(if a rain cheque lets you buy a $20 dinner for $10, you don't want to sell that rain cheque for $9)
At Interactive Brokers there is no fee for exercising options, so when I've occasionally found myself in that situation
I have exercised the calls and sold the stock, a bit at a time, which avoids the need for a decent bid in the options market.

* The best day to BUY a new position is when the price is low, obviously.
But the best day to ROLL is when the price is as high as possible and the market is calm. (and forecast interest rates low!)
That's because, when you're rolling, you are a net buyer of time value.
You are selling something with a small amount of time value
For any given strike price, the time value is lowest when the stock price is as high as possible relative to that strike, and when the market (and that stock) is calm.
A high price is good because the implied interest rate at any given strike is lower.
Or, looked at another way, for the same interest rate you can use a higher strike, meaning you are tying up less cash.

Often you can roll to a considerably higher strike and the roll transaction will free up cash, often permanently.
This is essentially how I feed myself. It doesn't happen every year, but it generates a lot of cash when it happens.
A portion of my portfolio can be thought of as a perpetual set of calls expiring 1-2 years later, with a strike price around $X lower than the current stock price.
I will stop this when the expected value growth rate from Berkshire minus the implied interest rate is not a big enough gap, causing the party to end.

So, overall, the ideal time to roll would be when interest rates are falling, on a day that the stock price is high,
at nearly the earliest date that the two-year options are available.
(not the first few days...liquidity seems to take a couple/few days to appear for a new contract)

Jim