5 ms·
Suppose you've made such a prediction about a tech company. How do you act on it? You can buy stock, of course, but is there a more powerful action available?
by presidentender 16y ago
Suppose you've made such a prediction about a tech company. How do you act on it? You can buy stock, of course, but is there a more powerful action available?
- tlholaday 16y agoSell the stock of its competitors short.
- mattmillr 16y agoDon't you have to borrow stock to sell it short? Can you do that for 6 years?
- aidenn0 16y agoTaking a short position in the competitors is troublesome because you need to time it right. If you think apple can win, you can go long any time the stock is low and end up a winner.
- borism 16y agowho was apple competitors in 2000s anyway? I don't recall anyone noticeable
- gjm11 16y ago"The market can remain irrational longer than you can remain solvent." -- John Maynard Keynes.
- francoisdevlin 16y agoWhat happens when the entire market segment balloons? APPL might beat out MSFT, but if everyone is more profitable, your short failed.
- zackattack 16y agoAre there publicly-available financial instruments that account for this?
- francoisdevlin 16y agoA put order would work well for this, IIRC.
- Timothee 16y agoSide-note: Apple's ticker symbol is AAPL, not APPL. It got me the first times too.
- rmc 16y agoYou have to know what their compeditors are. Pre iPhone, Apple's compeditors were Microsoft. The iPhone makes it look like Apple's compeditors now are Nokia/RIM (the makers of Blackberry). However a case could be made that Apple's current iPhone compeditor is or will be Google. So if you wanna bet on the iPhone, it's not clear who you should short.
- Tuna-Fish 16y agoLeverage and buy stock. The only thing that stops you from increasing the size of your bet is the risk of the downside. To some point, this is manageable, but after that point, it becomes pure insanity. YMMV. Shorting the competitors because you think a company is going to do very well is actually a bad idea -- it's entirely possible that Microsoft would have been able to open another very profitable market.
- barrkel 16y agoThere's a bigger problem. There's always going to be a spot-on prediction ex post. But you need to know which prediction is spot-on ex ante.
- vlad 16y agoIf you knew a technology was going to take off, learning their programming language/API and becoming a well-known expert on it could pay great dividends. Additionally, you could work there to make contacts and learn even more. Finally, you could try to join their developer programs while they are still in beta. Note that unlike suggestions to purchase stock in the company, this advice applies to technology at private companies, as well.
- zackattack 16y agoare humanized forms of indentured servitude legal anymore? for example, could you pay to send someone to the bignerdranch bootcamp to learn iphone dev, but then have some way of mitigating the risk that they don't then go ahead and auction their new skills, acquired at your expense, on the market?
- jimbokun 16y agoAren't these called non-competes?
- zackattack 16y agoNo, because I could train you to make iPhone apps for dentist appointments and then you could go sell your skills to make iPhone apps for farting. FartGenerator, Inc. is not in competition with iHaveAToothache LLC.
- nostrademons 16y agoThere are certain contracts you can make where you agree to work for X company for Y years if they pay for your education, or you have to pay them back for the cost. The U.S. Armed Forces does this with ROTC and the military academies. Several corporations will pay for an MBA if you come back to work for them afterwards, but you have to pony up the cash yourself if you choose not to work for them. In all these cases, the choice is always "Work for us or pay us back" - I don't think you can structure a contract so that they must work for you without running afoul of the 13th amendment.
- gxti 16y agoPrediction markets. They're like an elaborate, worldwide bet. Shame they're not government-sanctioned; a world with ubiquitous prediction markets would be interesting indeed. http://en.wikipedia.org/wiki/Prediction_market http://en.wikipedia.org/wiki/Prediction_market