It’s currently tax return time in Australia so this is a brief note on a couple of things I’ve realised lately. Note that these are Australian specific ideas that may or may not be helpful elsewhere.
Firstly, when selling shares, it makes life a lot easier at tax time if you match the quantity of each parcel sold with a parcel of the same size that you bought. This especially applies if some of your sales are inside the 12 month period where there is no capital gains tax discount and some aren’t. Splitting parcels up at tax time is not fun.
Secondly, it pays to have an easy way to include tax considerations in your buy decisions. I sell when the projected return for a stock falls below 11%, a concept I copied from Tony Hansen at EGP Capital. The reason I do this is that I don’t want to give myself brain damage trying to make sell decisions. I also think I have no way to know whether a stock is likely to go up or down in price once it gets that close to fair value. I set 11% as the hurdle rate that a stock needs to meet as that is slightly higher than the expected long-term return rate of 7% to 10% for most stock indices. If it doesn’t meet the hurdle rate I would be better off moving the capital from that holding to another one that does.
So far, all of the stocks I’ve sold have topped out at a maximum of about 15% above my sell point. One day I will sell something that continues to compound to a much higher price because of either momentum or because I underestimated the potential return. Neither of those prospects worry me at all.
When you hold a stock which has appreciated in price significantly, some of your equity in that holding belongs to the government in the form of deferred tax. It’s effectively a levered holding. When you sell, you need to pay out your debt provider (the government) and they don’t provide further funding until another holding increases in price. I want to compare the levered return on the stock I’m considering selling with the un-levered 11% hurdle rate so that I’m comparing returns in a fair manner when selling.
The annual projected return formula I used until recently is:
Projected return = (Dividends + Buybacks + (Retained Earnings x ROIC/Discount Rate) + Organic Earnings Growth + Discount to Fair Value/3 years) / Share Price
That assumes that a stock will re-rate to fair value over three years. You could also simplify this to:
Projected return = Running Return + Re-rating Return
My tweak to the above method is to substitute the share price in the denominator for the share price minus the capital gains tax per share incurred when selling.
You could spend a lot of time calculating the exact tax debt but you don’t need to. If you use a spreadsheet to keep track of your portfolio it is very easy to have one column for the projected return calculated the old way and one with the modified denominator using your maximum incremental tax rate and assuming no discount for holding more than one year.
This method is oversimplified but it is the worst case – I would only try and be more specific on the holding period discount and tax rate if there is a large difference between the columns for a particular holding. Most of the time the difference between the rough calculation and an exact calculation is rounding error in percentage return terms.
I’ve heard people comment that they don’t want to sell a stock that they’ve held for a long time which has appreciated in both price and value because of the large tax bill they would incur. Let’s use the above and see what the magnitude of the problem really is.
If we take the above formula, assume a fair value stock price of $1, a running return of 7% at fair value and a tax rate of 30% then we can come up with a formula that allows us to plot the sell price at 11% projected return versus the purchase price. I won’t include the formulas here (please email me if you want them) but the results look like this:
That doesn’t look like a very curved line but the sell price approaches about $0.89 as the purchase price heads towards zero. Another way to put it is that even if the stock is a 100 bagger, you still shouldn’t be waiting till above fair value to sell if you work on the same assumptions.
What happens if you think that your best alternative to the stock is cash and set your hurdle rate at say, 2%? To do this you would need to be extremely confident in your appraisal of the stock. I doubt I’d ever do this but here is what the curve looks like:
So under those assumptions you would be prepared to pay up to around 17% more than fair value in the unlikely event that the stock was a 100-bagger.
From the above I conclude that there are some conditions under which a typical investor should let a stock exceed fair value before selling it – but not by more than about 17%. That number will change depending on your tax situation – remembering it isn’t important.
What is important is to write out your own set of assumptions and reasoning regarding opportunity cost and stick to them. I don’t think you should ever wonder if selling a particular stock was a good decision. Better to check whether your return estimates and sell decision process are based on sound assumptions. If you want to use something like a momentum strategy to harvest price increases above fair value then that’s also fine but I would make sure to have a solid set of assumptions and a black & white decision making method for that process also.
Disclosure: This is not a recommendation to buy or sell any securities, nor is it financial advice. It’s definitely not tax advice. All information presented is believed to be reliable and is for information purposes only. Do heaps of your own research before buying or selling any security, especially any that I have discussed.