# Taxes: Selling vs. Holding

One of Saul’s strengths is his willingness to sell, especially to reallocate to something with more potential. As I’m in the midst of reallocating, that got me thinking about how to decide whether to sell or hold, and how taxes fit into that.

In a retirement account, there are no tax implications for selling. There’re just commissions. So reallocating to something you think will perform better is easy. But in a taxable account, it seems like you need to weigh the cost of taxes to decide whether to let a slower growing position continue, or take the tax hit and buy something growing faster.

For purposes of discussion, I’m going to assume the top Federal and California tax brackets for long-term capital gains. Federal is 20% (ignoring the Net Investment Income Tax). California doesn’t give a tax advantage to long-term gains, so you get the top 12.3% bracket, plus the 1% mental health services tax. That adds up to 33.3%. I’m going to pick two examples far apart on the spectrum.

First up: AAPL. I had a chunk of shares with a cost basis of about \$81 (post-split). I sold for \$96. That’s a gain of \$15. Income tax on that will be \$5, or just a tad over 5% of the sales proceeds. My new investment only needs to beat AAPL by 5% over some time period in order to be worth the reallocation. That’s sounds quite doable, so I sold and have been reallocating.

Next up: TSLA Jan 2015 \$30 calls (originally part of a synthetic long). Their cost basis is \$8.90. They’re currently worth somewhere around \$191-\$192. Let’s call it a gain of \$183. Taxes on that will be \$61, or just shy of 32% of the current value. If I sold to reallocate, the new investment would need to beat TSLA by 32 percentage points over some time period to be an overall win. That’s a more difficult hurdle.

How long a time period should I use for these comparisons? It should probably be a fraction of the average holding period, and Saul guessed his was about two years. One year seems an obvious and simple choice. I could imagine using a longer time period if you think both old and new investments had a long time horizon for growth, but I would not want to wait too long for the new investment to start beating the old one.

One argument is that I’d end up having to sell those at some point anyway (for retirement income), so I’d incur the taxes anyway. But at least the entire investment would continue growing until then. And if I’ve done my retirement planning and investing right, there will still be a huge nest egg left when we die, meaning no capital gains paid, and heirs get the shares at a new (higher) cost basis.

-Mark

P.S. For now, I’m selling covered calls against those TSLA calls. I’m selecting my strike price to get about 15-20% ARR if the stock stays below the strike, or about 18% return if the stock should surge too far past my strike to continue rolling. I’ll have more decisions to make as those \$30 calls approach expiration.

6 Likes

For purposes of discussion, I’m going to assume the top Federal and California tax brackets for long-term capital gains. Federal is 20% (ignoring the Net Investment Income Tax). California doesn’t give a tax advantage to long-term gains, so you get the top 12.3% bracket, plus the 1% mental health services tax. That adds up to 33.3%. I’m going to pick two examples far apart on the spectrum.

You can deduct state taxes from your Federal return at the nomial rate. Assuming 35% nominal, this brings the State tax down to 8.645% for a total of 28.6% (vs. 33.3%).

Mark,
I am with you
I have a synthetic long in TSLA, not at your level, but at \$160 for Jan 2016, so making some money and I have started selling a Call to collect premiums. If they get called away, so be it.

As for tax implications, what I have done in the past is sell my losers to offset my gains in order to limit the tax bite. If I like the stock that I sold, I wait 30 days to buy it back. If you are worried of the stock rallying within the 30 days, buy something comparable.
Good luck, Erik