I like to think of the investing discipline as composed of three key modes of operation.

Buying – Figuring out what you should buy and what you should not buy. There are many strategies that work here but my favorite is buying things that others are not buying. And my preferred reason for “others not buying” is that they don’t know about it yet.

Managing – This is the work an investor does to manage the investment. It includes decisions around whether to buy more of an investment that is working, which is incredibly important and can massively impact returns. But I believe that the most powerful thing an investor can do to impact their investment is to work with the management of the investment to make sure that the team is making the right strategic and operational decisions.

Selling – This is all about when to exit an investment and how. It is the hardest part of the investing discipline in my view.

The conventional wisdom on selling is that an investor should set a target price when they buy and once an investment reaches that target price they should sell.

That approach doesn’t work very well in venture capital because as a minority investors in an illquid investment, we don’t control the sell decision.

That doesn’t mean I/we don’t have targets when we make our investments. But it does mean that we don’t usually have the ability to do sell when those targets are hit.

And, as a result of this dynamic in venture capital, I have learned a different lesson over the years about selling and that is to let your winners run and sell everything else.

As I mentioned previously, as minority investors in illiquid investments we don’t control the sell decisions. They are made by the Board and driven strongly by the founders and management.

But that said, when we have the opportunity to sell an investment that is not one of our big winners, I have found that it is generally the right idea to do that.

When you make an investment that is really working out, I have found that it is generally a good idea to hold on to it even when it goes past your original sell targets for it. It can be a useful discipline to develop new sell targets when this happens based on the new information you have about this investment.

In the venture capital business, your best investments often go public and the venture capital fund distributes the stock to the underlying investors (called LPs) in the venture fund. Those investors then have to make the sell decisions on those investments.

As a general partner in these venture capital funds, I receive these distributions too.

And, as a result of some really poor decisions earier in my career around selling or not selling public stocks that were distributed to me, I have developed an approach for selling stock that is distributed to me.

I like to sell one third of the position immediately, put one third away for a long term hold, and actively manage the other third.

This method insures that if the whole thing blows up, at least we got something out. If it goes up 10x or more, at least we didn’t miss out on all of that. And it is simple and easy to execute and we do it this way all of the time.

But even with all of these lessons I have learned and approaches I have developed over the years, I continue to struggle with selling. It is hard for me to do and I resist the urge, particularly with the big winners. It is like taking your medicine. You know it is the right thing to do but it doesn’t feel very good when you do it.


Comments (Archived):

  1. Tom Labus

    To buy when others are despondently selling and to sell when others are avidly buying requires the greatest of fortitude and pays the greatest ultimate rewards. John Templeton Getting out has always been my Achilles’s Heel or hell in a lot of cases.

    1. Pointsandfigures

      yup-how many times I saw this in the trading pit. Fred’s post could apply to that sort of trading, although time was condensed. The mental gymnastics are the same

      1. fredwilson

        Gymnastics is a good word for it

    2. Girish Mehta

      The most important aspect of investing is the price you buy at, but the most difficult aspect is the selling.A bad purchase price makes a good company a bad investment, while a great purchase price makes a poor company a good investment.Buying is not difficult, but the buying price is most important. That it is easy makes us sometimes forget its importance(This is not a public quote, but apparently Buffett once said at a private dinner – Price is my due diligence).Selling, on the other hand, is both important and hard.If you get in a great price, the selling is easier (not easy). The likelihood you keep second-guessing yourself on the sale is higher when you got in on a not-great price.I don’t remember if it was Klarman or Graham who said – The purpose of the margin of safety is to make the forecast irrelevant.

      1. Tom Labus

        Templeton buying shares of companies at $1 in 1937.

      2. PhilipSugar

        A great purchase price makes a poor company a good investment.I would have to disagree with that. A great purchase price might make it easier for a good company to be a good investment.But unless you are in the business of scrapping out companies (and that is a business P/E people make a living at it but it does not work for startups) A poor company is a poor company is a poor company.

        1. LE

          Generally focus on whether a purchase makes sense and then try to get the best deal you can within reason. Don’t make a purchase decision if ‘I get $X more of a price reduction’. Exceptions of course. For example if you are simply trying to build in some runway then it makes sense to focus solely on the price. A small example is a piece of RE I bought (I hate to sound like I am JLM with this because I’m not in any way) but I bought it figuring that even if it was vacant for 2 years it was still a bargain. Maybe even three. But even then it would still be a bargain if the price was higher. So I just decided to buy it and then simply tried to get the price as low as I could knowing that I was going to buy it no matter what at the level it was at. As someone famous once said ‘all the risk had been taken out’.

          1. PhilipSugar

            Again you are talking about hard assets. Is your core business a hard asset? Sure you want to know about why a hard asset is distressed. Frankly if I was willing to take the brain damage I’d be a slum lord where 6 miles from where I live. You know I choose otherwise.Frankly I think your renting to people that have poor credit is a good deal. Same for Payday Loans. Just have to decide where/if you want to draw a line. Not debating morality.

          2. LE

            Frankly I think your renting to people that have poor credit is a good deal.It is my way of GIVING BACK.

        2. Girish Mehta

          Let me unpack that.1. Referring to public companies in response to @TomLabus:disqus comment. Not startups.2. “A poor company is a poor company is a poor company”.Agree. If you read closely, I am not saying different.3. “A great purchase price might make it easier for a good company to be a good investment”.Sure. But the thing with *mostly effiicient* public markets is that good companies are not available at a great purchase price.Finding a good company that is available for a great purchase price is another way of saying that one is smarter than the entire market (else the price of a good company would not be discounted to this great level). These opportunities exist, are rare, and are part of what value investing is about. But lately I have come to believe that value can also be found in a poor company that is trading at a great purchase price.We are familiar with the idea that – That company is doing great, but is it really worth 50-60 times earnings (late 90s).This is simply the inversion of that statement.If there is a too-high price at which a good company is not worth buying even if it remains a good company; there is a too-low price at which a poor company is worth buying even if it remains a poor company.The difference is not in buying but in holding. When you buy a good/great company, you continue to hold despite price movements unless your initial thesis is proven wrong. When you buy a poor company based on a great price, you need to act quickly once the price moves.

          1. PhilipSugar

            Since we are on AVC I was mainly talking about venture investments.I have seen VC’s invest at very low values. It doesn’t work out when the company flails and the founders and employees are bitter.”Catching a falling knife” is tough. I have also seen and lived where you think it can’t go lower and it does. It can’t trade lower than cash on hand…..except there are so many liabilities and you don’t have control of management that it keeps going down.I do not disagree except if it is a good company (not great) that you think is out of favor, maybe people were valuing like a great company and it had a bad couple of quarters because things didn’t materialize as fast as the quarterly mentality dictates. Then yes things over correct. I see this with many companies but especially retailers. Great ones have done great. Good ones have struggled but prices dip and come back. Poor ones? That train goes one way and it is down. I give you the poster child….Sears. I don’t know if I would want to ski down that chart

  2. awaldstein

    I love this post.And a believer that discipline over time is a winning strategy as well.Also true with how we handle our health and relationships also.Life if we are lucky is all about the long game.

  3. Rob Underwood

    Great post. Best part is that it doesn’t include any discussion of trying to time sales based on short and long term tax consequences (understanding that in your case it’s almost always a long sale). I still remember some cautionary tales about engineers during web 1.0, I think at Cisco, who executed options and then held shares hoping to be taxed on a long term basis, and then ended up with huge 6 figure tax bills and no gains for themselves. Taxes are a consideration in selling but I worry when I hear people making buy sell (and execute in the case of options) decisions based almost entirely on tax considerations.

    1. fredwilson

      i never factor taxes into my investing decisions. many do. there are comments in this thread suggesting tax strategies. but i don’t let taxes enter my head when making investment decisions

  4. jason wright

    i wonder how Lance Armstrong managed his Uber investment? JLM, do you ever see him knocking around in Austin?

  5. Jeff Hohner

    Timely post – thanks. Thank you as well for reconfirming my thoughts around this. It seems most great investors including Warren Buffett et al suffer from the ‘when to sell decision’. It is reassuring to know that I am in good company and am not alone. Letting the winners run and monitor is important – they are so hard to find/replace.

  6. Emi Gal

    Fred, what does “actively manage the other third” mean in this context?

    1. fredwilson

      Keep it front of mind. Watch how it trades. Look at earnings reports. Think about valuation. Always consider whether I should sell some

      1. Pointsandfigures

        Use option boxes to limit downside….

  7. Michael B. Aronson

    Great advice Fred. We are hopeful we will have first Ipo soon and I’m going to share this with lps when they ask me what to do. One thing that could impact the sell vs hold is if the stock is qsbs qualified for cap gains exemption, my understanding is that this stays with the shares after ipo so future gains are tax free

  8. Ari J. Greenberg

    What advice would you give to startup employees when their companies go public? Same advice or different?

    1. awaldstein

      Employees, especially management are subject to a host of other restrictions depending on their roles and their ownership stake.In my experience, rarely simply a matter of hold or sell though of course that is part of it.

    2. fredwilson

      More or less the same. But dependent on how much of their net worth is in the stock. The higher that number the more I would urge some diversification.

  9. falicon

    The rule of 33 is also a hard lesson learned in project and consulting work.No matter the client, how badly you want the project or how much you trust the client, get 33% at start, 33% at a defined mid point, and the rest at completion.Thinking more about it, the rule of 33 seems to be a universal rule on risk really.

    1. LE

      how much you trust the clientBack in the 90’s I paid $6000 to a brother of the woman who ran At the time she was a big deal. The guy had mental problems and totally collapsed and wasn’t able to provide anything at all. Zero. Everything about him checked out (including his at the time famous sister (she had written books and well this was the 90’s). For that matter he had written books (some with O’Reilly which as you know at that time was the bible of beginners). I think he was manic depressive. I just walked away so I wouldn’t get aggravated. [1]The first sign was when he started to tell me that he was going to build a tool to help him manage the project as a first step. So he was not focusing on the project but what I will call (in the spirit of ‘wash the suds off’) ‘Organize and get ready more fun than actual work’. Like buying the paint at Home Depot.Oh one other thing with trust. Someone coming with a referral from someone you know or trust means very little. The person referring is not going to pay the bill and back up the referral.Also one other thing. People who get referred business are not likely to go back to the person referring and say anything bad has happened. Why? If they do they might think that the person referring will stop sending business for fear of having it come back to them. So you sort of have to swallow it.[1] Thanks for your comment now I am relieving the aggravation!

      1. falicon

        :-)I’m here to help. ** assuming you have 33% of my helpers fee ready to transfer!

  10. Frank W. Miller

    I like your overall strategy sketch. The only thing I would add is that in a couple of highly appreciated stock gains I’ve had over time, I generally sell just enough to cover the taxes.

  11. Brian Lockhart

    Gah. I’m super guilty of this; I’ll cling to an investment / buy and hold long past whatever high target I had previously set. Even with substantial paper gains screaming at me to take them. I tell myself “hey, look at all those capital gains taxes you’re NOT paying!!!” 🙂

  12. PhilipSugar

    This is a great post. I would only add two things. As somebody who has sold two that were not big winners, and seen several people not sell and then get nothing. And also took a bath on another companies stock.Companies that are huge winners and by this I mean you KNOW you could get three offers instantly and your fundamentals are not going to change can be sold.The rest are bought not sold. Meaning you are not going to manufacture a buyer, the buyer will find you and when they do and you don’t sell that opportunity might not come along again. They very act of trying to find a buyer actually reduces your desirability. If the buyer thinks they found you, you can be a princess, if you come looking they will assume you are a frog.As for taking stock. The list is long of people that I know that took stock and weren’t immediately able to execute Fred’s strategies due to lockup that have lost.

    1. LE

      The rest are bought not sold. Meaning you are not going to manufacture a buyer, the buyer will find you and when they do and you don’t sell that opportunity might not come along again. They very act of trying to find a buyer actually reduces your desirability. If the buyer thinks they found you, you can be a princess, if you come looking they will assume you are a frog.This is so true. Can’t agree enough. Totally relevant to what I do actually.One thing that is big that factors into the person with the ‘merchandise’ though is both how much they need the money and also to what degree they are a gambler and finally how rational they are. The best is a rational seller that needs the money. Then it goes down from there to the rock bottom which is ‘gambler who is not rational and doesn’t need the money’.Of course there could also be cases where someone who doesn’t need the money ends up being a good buyer or seller. In this case emotions need to be manipulated. And that can be done in many cases. You deal with ‘nuts’ so not always.The example I will give is my father and the girl who wanted to buy a religious menorah from him at a gift show he was an exhibitor at. (He was a wholesaler but agreed to do a retail show one time).This young girl walks up and wants a menorah. My dad tells her the price let’s say it’s $25. The girl makes a counter offer and triggers my Dad’s negotiation reflex. My Dad who doesn’t need the money turns her down and doesn’t move. Sorry he says.Well the girl ends pacing back and forth for the next half hour in front of his booth. She telegraphs (and this is important) “I want what you have but I can only afford to pay so much for it”. So she doesn’t walk away (wouldn’t work with my Dad he doesn’t care, right?) she actually acts more interested in what she can’t afford. Finally my Dad goes out and tells her ‘ok I will sell it to you for $x’. He felt bad for her I am guessing. She gave him a reason to do the deal that transcended the money, the negotiating reflex, the gambler (he was not a gambler but same would apply) and so on. In the end both parties were happy. He felt good I will guess got the ‘party in the brain’ thing going thinking he was a good guy and had helped this girl out.Anyway very important lesson. It’s not always ‘your merchandise isn’t worth it! You are nuts!’ sometimes it’s ‘I love it, the price is fair, but honestly I only can pay $x pretty please help me out’.This will not work when buying office buildings of course (but parts of it will work when dealing with people selling office buildings). It’s all psychology. Deals are people.

      1. PhilipSugar

        Here is what people don’t understand, they think you put a company up for sale and then you get buyers, like a car, boat, real estate, etc.Answer: NoExceptions: You might have pull where somebody can get buyers because they know if they don’t they will miss out on other deals. Fred is an example, a division of a big company might be the same, there are others.Big exception: Somebody that finds you and wants what you have.But you as a non rock star company?? The only people that will make money are the business brokers that will try and charge you a retainer fee. Might as well be the patent companies you see on TV.

        1. LE

          Might as well be the patent companies you see on TV.My sister’s first husband (in the 80’s) had this stupid idea my Dad couldn’t stop making fun of him. To his face yep that’s the way you do it. It was called the ‘refresherator’ and the idea was if you were watching football you could grab a beer w/o getting your lazy ass up from the couch. Not only did my dad not watch sports but he didn’t drink beer and he wasn’t lazy. He just thought the idea was so stupid. And he would tell it to his son in law. Anyway the son in law would say ‘yeah but the lawyer thinks it is a good idea’ (or the patent consultant for that matter). My god how could someone be so naive and stupid to not see why?Of course he never did the idea. Years later some big company made a similar product. But the patent guy telling him it was a good idea was the cornerstone of those ads on tv that you speak of (I think it’s called ‘inventhelp’). You now my ‘if the woman wants a mink coat you sell her a mink coat you don’t tell her a cloth coat will work and is cheaper’.One thing I will say though. Hopes and dreams are worth some money. You know that does have value in some way I will say that. Honestly no more stupid than many other things people do which lead to nothing.

        2. LE

          car, boat, real estateWill add ‘airplane’ to that list.All those things have a few things in common.For one thing this is an active wholesale market for the product that establishes a liquid and more importantly a price bottom that can support a decision not to sell when you get an offer.For a company there is no such thing. Ok maybe for liquor stores or convenience stores or dental practices or even medical practices there is an active and ready market. But not for a typical business (ok add pizza shop to the list and I am sure there are others). When someone came and bought my company years ago I got exactly one offer and had to decide on that. Pass and would probably never sell it for anything close. That is what you are talking about. (Ok only other offer was from a big operator for 1/3 what I sold for).For the things that I sell (which I will not mention but you know what they are) there are two types of merchandise let’s say.One is merchandise where there is an active wholesale market that has established a bottom price for the merchandise that I can usually get if I needed to sell for some reason. Then there is other merchandise where there is no active market at all no price bottom. It’s venous fly trap. You get your one shot to sell and you rationally decide if it pays to wait for door number two. There is no wholesale market at least not at a price that it would pay you to ever sell at.

          1. PhilipSugar

            Yes, cash flow business like the ones you are talking about yes.

    2. Pointsandfigures

      And when you sell never play woulda could a shoulda. Don’t look back.

  13. JLM

    .On the sale side, a way to prevent post-sale erosion of a public stock is to include a “value savings clause” in the deal.A value savings clause says that if the stock price dips on the expiration date of the lock up or the anniversary of the transaction, the seller issues more stock to bridge the dip.I did this when selling a portfolio of office buildings to a public REIT. Everybody looked at me like I was nuts when I had it put in the docs, but I got it.It turned out to be an very important decision on my part as the stock price dipped and they issued more shares on each anniversary for 5 years. What was really important is that REITs pay a meaningful dividend and I was getting more stock which was keeping me whole on the original deal, but I got a lot more dividend paying stock.After I did this, I saw it being used on a lot of similar deals. It is not appropriate all the time because of the structure of the deal.JLMwww.themusingsofthebigredca…

    1. sigmaalgebra

      WOW! Actually CAN do such a thing!!! A keeper! Thx.

    2. LE

      A value savings clause says that if the stock price dips on the expiration date of the lock up or the anniversary of the transaction, the seller issues more stock to bridge the dip.Timing of that is critical though. Usually with that type of thing (because honestly if I understand your point why would they agree and/or not adjust their offer?) it has to be presented in reply to a last offer. So it is what I call ‘kickback’. That is you always need to make the other party feel as if they didn’t leave money on the table. So you throw in some extra term (usually nominal) so they feel they are winning.So it is presented as ‘ok I will do that ‘if”. Car dealers do a version of this (not the same thing but way to illustrate) with ‘ok but only if you give me a deposit today.’Point being they make you think you have made a good deal (among of course not allowing an open offer obviously and getting you to sign and buy).Presentation is critical obviously. If you do bad acting they will know what you are doing.Everybody looked at me like I was nuts when I had it put in the docs, but I got it.I am curious who the ‘everybody’ is? It’s not nuts at all.The best is when I run into attorneys who end up saying things like ‘oh they won’t agree to that’ like thanks for making their job easier idiot. Golden rule is don’t ever worry about what someone else won’t agree to that is their job.

      1. sigmaalgebra

        Price negotiation:For my first new car, I shopped carefully and decided just what I wanted. It would have to be ordered, and I’d have to wait about six weeks.I didn’t know what a good price would be. So, I picked a price likely too low and offered it to the dealer closest to my home. When they quit laughing, I drove to several more dealers who could also order the car and offered the price to each of them. They all laughed.I raised my offering price a little and offered again. At one dealership the salesman saw what I was doing, got angry, and just turned and walked away. Hmm — looked like I was on to something good!I kept up this little iteration. A few more iterations, and at the dealer closest to my home the salesman went to the sales manager and we all sat together. The sales manager added to my last offer a small fraction of my step size and said that that was his “profit”. At that point, it wasn’t worth my completing another iteration, and we closed the deal.It was a pretty good car — small body, big engine, no power steering, brakes, or windows, no A/C, no radio, relatively simple mechanically, 4 speed stick shift, etc.Problem was, the rear axle ratio was about 3.23, and the first gear was about 2.52:1 — total bummer since 4th was only 1:1 with no overdrive. So, it took 3000+ RPM to cruise at 70 MPH to Indiana to see my fiance; I usually drove faster than 70 MPH, and I commonly got only 12 MPG. With better gearing I could have had still better 0-60 MPH and 1/4 mile elapsed times and high speed highway MPG nicely over 20.Just needed rear axle ratio of 2.56 or 2.29 and much wider gear ratios in the 4 speed transmission, e.g., first gear maybe 3.5 or 4.0.Since then, some high performance cars, e.g., Corvette, have such gearing, in that case, via two levels of overdrive, and can get MPG well over 20.It looks like the engineers wanted to keep down the maximum torque on the driveshaft while raising the RPM. Okay.

        1. Lawrence Brass

          Must be tough to sell a car to a mathematician, only second to sell a car to LE. :)I need both of you to help me buy an old Volvo XC90.

      2. JLM

        .In this instance, we had been trolling this portfolio around on the QT testing the market. The buyer needed Texas exposure as part of their strategy to do a secondary offering. The secondary was already in the red herring stage, so we knew it was for real.We were the only game in town (Texas) at that instant in time.When you are doing a deal like this, there is no protracted negotiation. You offer up your desired contract. The other side’s lawyers tell you you are nuts – cause they wanted to write the contract.In this instance, they were associated with Sam Zell and another big institutional buyer, so they were used to tough contracts, but they wanted to pay with cash and stock. Most sellers didn’t want the stock, but I not only was willing to accept it, but I wanted it because of the enormous dividend.The trade was our taking stock for the value savings clause. They were only recently public but owned by lots of big institutions.You have to have a commercially reasonable contract which provides for a real due diligence period and a data warehouse wherein you provide every lease, estoppel certificates, financial records, vendor files, tax bills, utility bills, maintenance records and “as built” drawing.You have the data plate from every piece of equipment and a list of all the critical elements – elevators, windows, roof, central plant, etc.You have to prepare to sell the properties so that the buyer can get in, do their due diligence, make an offer. Once they make an offer, you negotiate the numbers, but you have already laid out the contract, so you don’t lose time in getting on with it.Selling at that level is like preparing for an audit, you do as much as you can before the fact so once you agree to the numbers, you can just fill them in and the due diligence period starts.Obviously, the title work and surveys are also done.Like any deal, they tried to chisel us at the end. That is a funny story.When you are dealing with professionals there is really not much negotiation other than saying “I want” and their saying, “I offer.”To get your price, you have to inject real competition into the deal. I used to tell people I wanted the price with the biggest arithmetic mistake in it.JLMwww.themusingsofthebigredca…

        1. LE

          You have to have a commercially reasonable contract which provides for a real due diligence period and a data warehouse wherein you provide every lease, estoppel certificates, financial records, vendor files, tax bills, utilitiy bills, maintenance records and “as built” drawing. You have the data plate from every piece of equipment and a list of all the critical elements – elevators, windows, roof, central plant, etc.Who typically performs this function for the buyer? In house staff? Hire outside firm (accountant or consultant)? It depends? If outside what are the fees that this costs as fixed or percentage of the deal? More importantly what liability does the outside party have for errors omissions or mistakes? (And of course you verify they actually have coverage in advance I assume if you are a buyer I mean I would if it was my money on the line).All this comes into play with small deals. My variation is ‘HVAC guy climb on roof and shoot picture of mechanical system plate so I can see the age (in other words I don’t even trust what he says and I am not climbing on any roofs). When I colocated equipment I wanted to inspect the generators for the carrier hotel even though it was used by AT&T and big players. Easy enough to not trust and verify first hand.

          1. JLM

            .Typically a buyer promulgates the contract. Most sophisticated buyers have a standard contract they like to use. Institutional buyers will have a contract that is “approved” and debt driven buyers will have to have their contract approved by their lender.I did so much business with GE Capital, we had a contract we both liked and they adopted it for all of their portfolio.For guys like me who had been in the biz for a long time, I had an approved contract, an approved Letter of Intent, and a clause to address every potentially contentious point.Somehow, legal fees always end up being 0.5% of deal size. In Tx you used to get a 40% rebate on the title insurance premium which went to your lawyer who credited it against your account. Now, this is not prevalent because of conflict of interest considerations.Everybody involved has E & O insurance. If a pro has made a mistake, it has to rise to a level of materiality. In 40 years in the biz, I never had a problem with a professional, but had a couple of title issues which the title company fixed.As to equipment, I had a rule of thumb – HVAC equipment was only good for 7 years. I would have a guy look at all the data plates – pictures were mandatory – and then test them. Anything older than 7 years, I asked for the seller to give me a credit for a replacement at an agreed upon $ cost/ton.This is a huge deal with apartments – carpet (5 years), appliances (6 years), and HVAC (7 years). I have bought and sold more than 30K apartments.When I had my guys operating the property, I always felt like I could get 150% of target life, but it cost money to do it. I was also pretty tough on recovering tenant damage to carpets, in particular. Why I liked women tenants, they are not very hard on the physical plant. Hated puppies who could destroy a carpet in a year.I would have a matrix made which would show the equipment, the brand (some brands I didn’t accept), the tonnage, the age, the current readings (to detect leaks) and then we calculated the total cost at ZERO years and the remaining useful life at the actual age.I would get a deduct in the price for bringing everything back to zero.If you have a real reason why you want something, you often get it. Sometimes, I was buying stuff from institutions so cheap that I would almost feel embarrassed asking for an equipment deduct.In the real estate business, there are people who are reliable. But, when buying and selling, never trust anybody including yourself. Get the due diligence correct and sweat the details like the HVAC matrix.You have to turn RE into a totally numbers business to generate the required and desired returns.Apartments, in particular, get away from you a penny a SF at a time.JLMwww.themusingsofthebigredca…

          2. LE

            You have to turn RE into a totally numbers business to generate the required and desired returns.100% correct for most of business. [1] (And read my comment elsewhere about Fred and emotion ie ‘have a hard time selling’.)[1] For sure cases where emotion is valuable but probably more cases where it can end up having you end up on the wrong side of a transaction. (Or in a relationship with dating).

  14. Salt Shaker

    Selling stock at a gain is easy, selling at a loss is hard. Swallow hard and move on.

    1. JamesHRH

      Selling is hard period, up or down.

      1. Salt Shaker

        To each his own. I never have a prob unloading at a reasonable profit, even when a stock subsequently rises. Disheartening yes, but can’t complain too much if still a solid investment/return. AAPL for me is a case in point. Made a nice gain, missed the run to $200+. My biggest challenge is selling. If I have strong convictions about a company, then I’ll double down as the stock declines. I’ve been more apt to do that than unload (at least short turn). Did get burned on Nvidia, though. Cashed out before big run.

    2. Pointsandfigures

      Selling too fast is easy too.

  15. JLM

    .One of the easiest ways to take a loss or book a gain is at year end under the banner of tax selling to minimize your tax bill. Under the “wash sale rule” you can buy back what you liked 30 days later.It is like aspirin.JLMwww.themusingsofthebigredca…

    1. sigmaalgebra

      I continue to guess that businesses and investors could use some help in what decisions to make over time. For just how to do this, in surprisingly general situations, the math is now old, fairly well documented, and rock solid.There are some problems:(1) If not careful, the computing resource requirements explode, as in take over all of Amazon’s servers for the week — literally. The first solution here is to attack only relatively simple problems, and there should still be plenty of problems to attack and do well with. Then, as the computing resources needed explode as the problem sizes increase, on the flip side the computing resources needed contract at the same rate as problem sizes decrease.(2) Very much would like a lot more good data, e.g., on how to project interest rates, inflation rates, etc. Often can’t get the data really want, and even when in principle the data is available there can be a lot of work in finding, acquisition, cleanup, etc.(3) Get and publicize some successes so that the target users can make use of the capability.One of the maybe strong motives for doing this is tax law details such as you mentioned.For what can look like to a business person, imagine doing planning for, say, five years, 60 months. So, have a spreadsheet with one column for each month, time increasing left to right. Have one row for each important quantity — cash in the bank, sales, expenses, interest rates, inflation rates, foreign exchange rates, and various decisions on taxes, hiring, leasing, new projects, etc.Then at each month, fill in the cell values, e.g., get the current bank balance, from the previous month and that data seen there and decisions made there.Then ask for the decisions that maximize the expected value of the company at the end.A crude approach would be to determine fixed decisions — like in football calling the next four plays all once and not changing. Instead, the plays are called using the information available from the results of the last play. So, on first and ten, there is quite a collection of possible futures, or future scenarios or sample paths, where the decisions change at each play. Yes, also in this context the math says what to do, to maximize the expected results, and where no means of making decisions in the context can do better.Since my Ph.D. dissertation was in this area, I know something about it. For that work, I was able to make the computing needed quite reasonable.In recent years, the Department of Operations Research and Financial Engineering at Princeton has taken this problem seriously. They have some approximation techniques and compare with what would the best results be if we knew the future randomness, e.g., interest rates, exactly. That’s cute: Get free to do a lot of approximating and, still, have a figure of merit to tell if in the end the approximations helped or hurt!For just a few decision variables, say, hopefully as many as a dozen, commonly the computing resources needed should be small enough for current computing.Curiously, some years ago D. Bertsekas at MIT saw an interesting application of neural network empirical curve fitting as part of approximations that could reduce the computational resources needed.Long R. T. Rockafellar at U. Washington had USAF funding for research on this problem. Other big names in the field include R. Bellman and E. Dynkin.This project is sitting there, and some progress is likely available. But I’m not going to pursue it — my startup is easier to do and potentially much more valuable!Still the need is there; it’s a shame that people still have to make such financial planning just by intuitive and back of the envelope techniques.

  16. JamesHRH

    The answer isn’t three, but the good answers always come in a set of 3s.Selling is against your nature, you are a hoarder!Great post.Any fool can buy stock, the tough part is selling.Hell, this rule applies to every big ticket item: house, car, boat, etc. Most people focus on their ability to buy luxury items…..but your ability to sell them lessens your risk.

  17. sigmaalgebra

    For But I believe that the most powerful thing an investor can do to impact their investment is to work with the management of the investment to make sure that the team is making the right strategic and operational decisions.Perhaps in some definite sense, stockholders and the BoD have a right to give such advice with the CEO listening. Perhaps.But there’s a lot of advice out there for business on “the right strategic and operational decisions”, so much that a CEO can feel swamped. Since so much of the advice is different, “the right … decisions”, if present in the advice at all, are at best a small fraction of all the advice. So, the CEO needs means to evaluate and filter the advice.I know; I know; 90% of being a CEO is each of (i) selling, (ii) fund raising, (iii) making “the right … decisions”, (iv) building the team, (v) being a leader in crucial new technology, (vi) motivating the team, (vii) keeping up good PR, (viii) budgeting and making sure the bank account stays nicely positive, at least (i)-(viii). Hmm — let’s see, 8 * 9 = 72 so that we have work enough here for 7.2 CEOs. Poor guy won’t get any sleep; his kids won’t know him; his marriage will fail; he will be too tired to do his job well; he won’t get exercise or pay attention to his health; he will die and leave the company in shambles!!!!!Sometimes on some topics some VCs have good experience and advice. Sometimes.And there are other credible sources of advice on common challenges in business. E.g., getting good terms on leasing space for offices and the server farm, when the business is growing quickly. E.g., in hiring, have a technical training program or not?As a technical founder of a startup where some original applied math based on some advanced pure math prerequisites is the crucial core, enabling technology, technological advantage, intellectual property, Buffett moat, sure, I’d like some advice on how to improve the math!!For more? Sure: How to do better in server farm monitoring for, diagnosing, and correcting problems in performance and security.

  18. sigmaalgebra

    A lot of this thinking on buying and selling in business is to yield some psychological feel good conditions, that is, attempting to maximize a utility function that gives a lot of value to feelings and is not the same as just maximizing expected value in dollars.E.g., lots of people play the lottery where maybe they feel good; still, clearly the expected value in dollars is negative.When making investments over a long time in many situations, what will dominate is just to maximize the expected return, that is, use a simple utility function.For that maximization, for the limited data commonly available, for a first cut might want to build a model that fits the data well and otherwise is credible and use that to make the decisions. As usual, if limit the input data to own experience, then can lower bias but suffer a large variance; if use a lot of data also from the experience of others, the market, etc., can increase bias but lower variance. Long term, it’s better to get the bias low and put up with the larger variance instead of getting a higher bias but lower variance.Uh, for this model building, I wouldn’t want to insult and denigrate it with nonsense such as data science, machine learning, artificial intelligence, each a contradiction in terms and a fad with a really ugly history!IIRC, utility function theory goes back to von Neumann and has a nice axiomatic basis.

  19. Advice from a small VC cowboy

    For small VC cowboys your advice in sensible; for large VC farmers and large VC city dwellers it is not.Your advice inherently presupposes that VCs have dearth of time and scarcity of skilled labor similar to that associated with a land grab such as one that might occur during a gold rush (starting around 1848-1849 during The California Gold Rush).VC farmers and VC city dwellers have plenty of time and plenty of skilled labor. Also, large VC farmers and VC city dwellers won’t normally be tiny minority stake holders. Even when they do not control a majority interest of an investment, they will generally be relatively far more powerful than a small (boutique) VC firm like Union Square Ventures.It is easy to imagine investors in on a deal saying “Who cares if we upset the guys at Union Square Ventures? They own about 4% of this company. If they don’t like our decision, too bad for them. After all, they are small and weak”On the other hands, it is easy to imagine investors in on a deal saying “We had better not upset the guys at Goldman Sachs? Although they only own about 15% of this company, we are in on six other deals with them. Also, if we upset them they might retaliate against us by freezing us out of other deals or attempting to harm us in another way. Those guys are big and powerful. Therefore, we need to treat them with kid gloves.”It’s the law of the jungle: the large and powerful dominate the small and weak.

    1. Yichen Zhang

      From a single investment perspective, smaller holding is easier to trade. Thinking too much about whether upset management and other shareholders reduces flexibility, an advantage of being small.

  20. creative group


  21. LE

    But even with all of these lessons I have learned and approaches I have developed over the years, I continue to struggle with selling. It is hard for me to do and I resist the urge, particularly with the big winners.Yeah you have to get over that reflex. [1] Not a good thing for someone in business in which the idea is to make money. Primarily.I have noted for example what you have said with regards to both art and selling crypto. You seem to have some kind of fixation and reasons that guide you that maybe shouldn’t. Maybe it makes sense with art (because I assume you aren’t buying it to profit although you must feel it has value) but with crypto honestly it seems counter productive.A good exercise (that nobody ever does) is flip the situation. If you could sell something for $X would you buy it for $X today if you didn’t own it and had the money. The answer to that is almost always ‘no’ but what people already own they will hold onto until (and this is important) ‘the bitter end’.[1] I am lucky I will completely fall apart with anything medical related (meaning even if non-serious) however business wise, selling, negotiating, buying I am able to be 100% rational and therefore am suited to make a decision which is not based on emotion. That has been tremendously helpful I have found.

  22. Pointsandfigures

    Reflecting on this post, I wanted to spin it a bit differently if it’s okay with Fred. Fred outlines a case where you take a position in a company or an investment after doing a lot of research and figuring out that you have asymmetric information. You invest, and continue to press it if it’s a winner because you have more information than the market and you are taking advantage of it. That’s smart investing.There are times when you are just lucky. In the trading pits we’d call them “out trades”. (by the way, if I ever buy a boat it will be an out trade because I am not a mariner and hate boats) An out trade happened because you thought you were buying or selling and the other person thought they were doing the same thing. Or, there could be a price differential or a small error. Buy/Buy or Sell/Sell were the worst kinds of out trades because if you were wrong, you had to make the other person good and then re-buy or re-sell in the market which was almost certainly going to be a worse price than you thought you were doing the trade in the first place. Occasionally, it would be a winner and it was like drawing the Monopoly card, “Bank error in your favor”. I knew people that had winning out trades and all of a sudden thought they were smart and traded out of it and lost it all. All they had to do was pull the trigger and cover and they made some money.Then there are other times when a wave happens. You are fortunate enough to be a part of the wave but you didn’t get in on the ground floor. This is crypto. There are people that bought Bitcoin for example because they heard about it from someone and took a flyer. Now it’s run up in price and all of a sudden they are telling themselves how smart they are. Of course they saw it was new tech and was going to change the world. The reality is they got lucky and caught a wave and the only course of action is to get out before it hits the beach.There are some maxims in trading a position that never change. Pigs get slaughtered is one. Knowing when you got lucky and when you had skill is another. It is often very very hard to tell the difference because we are human.

    1. LE

      Now it’s run up in price and all of a sudden they are telling themselves how smart they are.Problem is over time like the casino if they follow this strategy the house will eventually take all of their money. That said it’s a personal peeve of mine when people do stupid things and make great amounts of money.Pigs get slaughtered is one.The hogs get slaughtered, no? (The pigs get fat).

      1. Pointsandfigures

        Ha. If we want to get Technical Gilts and pigs get slaughterd. Sows only do when they are old or the price drops. I try never to mess with pit logic. It was pretty spot on and the principles malleable to a lot of different occupations.

    2. Tom Labus

      Knowing you’re lucky!!! Ball game.

  23. Pointsandfigures

    Do you have tomorrow’s copy of the Wall Street Journal?

  24. jason wright

    didn’t you dump all of your bitcoin?

    1. fredwilson

      no. where did you get that idea?

  25. Jose Paul Martin

    Fred, selling has been one of the most difficult things for investors, private equity, venture capital or public equity. I’ve faced situations where I begged to sell companies, while riding the wave and hitting the boom. Those companies are today worth a fraction of the original investment, factor in the time factor for recovery to breakeven and seeing a nice IRR is out of question. So the adage, “let your winners ride, cut your losers short” is pretty much good in public equity or stock markets with a shorter time frame, with longer ones… it might not be the case.

  26. Fraser

    I have some assumptions, but I’d value hearing your thoughts on why this is the case:”when we have the opportunity to sell an investment that is not one of our big winners, I have found that it is generally the right idea to do that.”

    1. fredwilson

      investments are like athletes. some are just way better at what they do than everyone else. and every once in a while we find one of them and they produce massive returns for everyone in and around generally takes three to five years for us to know which ones are those companies, but once we can identify them, it makes sense to hold onto them and let everything else go.but, as i said, it doesn’t always work out that way in practice. we still have two portfolio companies left in our 2004 fund and neither of those companies will be a top performer in that fund.

      1. Fraser

        Thank you

  27. Michael Rodov

    Schmuck insurance

  28. george

    Some people set and follow financial policies when deciding to sell, a disciplined approach. I’ve changed over the years, moved towards weighting my convictions at different stages of the business lifecycle – perhaps this is more concentrated but it’s been working out pretty good over the past decade.

  29. Hunter

    “the venture capital fund distributes the stock to the underlying investors (called LPs)”@fredwilson:disqus – Is this distribution to LPs always the case, or does it vary depending on the individual fund structure? I ask because I have been tracking the CFIU / FIRRMA issue and this seems highly relevant in terms of which firms will be impacted:

  30. Kristiyan Uzunchev

    First of all, great post Fred!I am in sales for a long time, read so many books that now selling is an inner sense.Just wanted to share with all of you guys, looking for leads – use Instagram!It works like a charm for me! –… 1.Create a trustworthy profile (try this checklist… ) 2.Find your potential clients ( use hashtags)3. Start sending DM’s implementing everything above that Fred wroteGood luck to all!