 Myself and also Chris Matz and Olaf Nassan. I come from an oil and gas background. Oil and gas have a lot of uncertainty. Chris and Olaf have a lot of experience in the financial sector and they've done a lot of work in understanding uncertainty in those areas. So what we're going to talk about today is what can we learn in the software world, some of the things that software experiences that are similar to what we see in other businesses. So first let's start with risk and uncertainty and some of the definitions that at various times you'll see relative to risk and uncertainty. Here are three different variants of definitions. One was from Frank Knight who was an academic and distinguished between called risk immeasurable and uncertainty was quantifiable. You see this in academic publications quite often. I don't quite understand how it came about or how it came to be used but I don't particularly like it because I don't see it consistent with the English language. So I'm going to try to not go that direction. The PMI generally has risk, understands that risk is there and says it can be the positive or negative risk which is again counter to the English language. So when I talk about risk and uncertainty in general I'm going to be looking at risk as a situation that involves exposure to danger like getting run by a bull or uncertainty is the state of being uncertain, not known or established or questionable. So that's sort of the basis that we're starting from. So since this is about risk I called it risky business. Risky business is a take off on a film that was made a few I think in the 1980s with Tom Cruise. I just want to show this little short clip here. We don't have sound. Can we check sound on the computer? The first thing about risk management you got to know is when you got too much risk. We're about to go there. The next step, I live in Houston, Texas. Houston, Texas is known for having real challenges with hurricanes and so this is an actual plot of the results of projections that I think 16 or 18 different weather services had regarding a path of Hurricane Rita. Hurricane Rita had come just after Hurricane Katrina had made a significant impact in New Orleans. It was a huge, huge devastation in New Orleans so Hurricane Rita was a Category 5 hurricane at the time. So it was a serious thing because had it come through and hit Houston as a Category 5 would be a big issue. So the interesting thing, what do we know about every single one of these projections? They're going to hit? Anyone else have anything to say about every one of the projections? The profile changes over time. Each one of them does change over time. Anything else about each one of the projections? Confidence levels for each? Okay, possibly. So the answer I'm going for, which is an interesting thing because I did this for high school kids and I got a different answer and they were smarter than me. My answer was that all of them are wrong because I'm an old guy and I want to be negative about it. The high schooler said every one of them is a possibility. That's the same answer but a much more fun way to say it. Everyone's a possibility, I say everyone's wrong, but that's the whole thing. In one sense, everything's wrong because if we were to hold accountable to the single projection that you had, it would actually be wrong. There's no way to be possibly right. The one thing we do know is something we can measure and that's the plus down there because that's pretty accurate to be measured. And it turned out that the actual path of the hurricane wasn't like any of them. But what we also know is that the collection of estimates was actually useful because it gave us a band, something we call the cone of uncertainty. It gave us a band of where we might be concerned about hurricane hitting. Again, then looking at that cone of uncertainty, by looking at this over a sequence of time, we can actually see and make decisions based upon where we need to evacuate perhaps and over time eventually it comes through. This is the same type of management and bounding of uncertainty and learning through things that you can measure. Some other examples of uncertainty. I love this one, a war example. So this is a quote. They couldn't hit an elephant at this dist. It was the last words of General John B. Sedwick, the Union Army Civil War officer that he uttered during the Battle of Spotsylvania in 1864. He was bragging that they couldn't hit him. And they did. I played this the other day. I'm going to play it again here. I think it's really a brilliant description of the importance of feedback. So this is another war example here of Gordon the guided missile. Gordon the guided missile sets off in pursuit of its target. It immediately sends out signals to discover if it's on course to hit that target and the signals come back. Now you are not on course, so change it up a bit and slightly to the left. And Gordon changes course as instructed and then rational little creature that he is, he sends out another signal. Am I on course now? And back comes the answer. No. But if you adjust your present course a little bit, a little bit further up and a little bit further to the left, then you will be. So he adjusts his course again and sends out another request for information. And back comes the answer. No, Gordon, you're still very wrong. You must come down a bit and a foot to the right. And the guided missile, its rationality and persistence, a listen to us all, goes on and on making mistakes and on and on listening to the feedback and on and on correcting its behavior in the light of that feedback until it blows up the nasty enemy thing. Then we applaud the missile for its skill. And now if some critic says, why it made a lot of mistakes on the way, we reply, yes, but that didn't matter. Did it? It got there in the end. All its mistakes were little ones in the sense that they could be immediately corrected. And as a result of making hundreds of mistakes, eventually, the missile succeeded in avoiding the one mistake which would really have mattered, missing the time. So the movie industry, all movies they aimed to be just as great as the Titanic. The Titanic was until recently the number one grossing film of all time, supplanted by Avatar. Unfortunately, most movies end up like the Titanic. 78% of films lose money. And only 6% of films contribute to making 80% of the profit. Huge area of uncertainty. I think this is an industry that doesn't necessarily knew how to manage uncertainty because it's so driven by egos and everything else because I don't think it's necessarily the most profitable industry overall. But there are other industries that are able to manage risk and be effectively very profitable. Book industry is a good example. It's my book. In 2004, out of 300,000 books that were printed or that were available for sale, less than 25% sold more than 100 copies. Isn't that amazing? Poker. Any poker players here? Texas hold them? All right. So just go with me here. These are three different possible hands. Which of these three hands do you think would be the best Texas hold them whole hand? Would you think it would be the ace four? The pair of threes? Or the six-nine? How many for A? You don't have to know a lot here. A couple of A's. How many for B? Would be more for B. How many for C? A couple more for C. If it was just those two cards and you were going to stop, obviously the pair would be the highest. Let's see what it has. 29.6% chance of winning. The ace four has a 33.5% chance of winning. And what looks like really crappy hand, the six-nine, actually has a 36.5% chance of winning. What's going on here? What's going on? In this case with poker, that your strength of your hand going into this is largely determined by your weakest card. Your weakest second card. So the stronger the weak card is, the better your chance of winning. We don't think that way. I'd like you to think about it also on your team. How much of your team's success is driven by your weakest link? We tend to focus on strengths. We tend to forget about what we need to do to bring our weakest links up. Sometimes you can bring your weakest link up and make a huge difference. Again on poker, I'm thinking of creating a metric for poker called percent of hands won. Think percent of hands won would be a good metric for poker? Success? How many think it's a good metric for poker success? Percent of hands won. What's that? I might not tell. It's a good point, very good point there. I was going to say though, it turns out that this is pretty good metric. It's a pretty good metric for determining if you're going to lose. Because the more hands you won, the more likely you're going to lose. Why is that? The more money you're likely to lose. Why do you think that is? You're playing too much. In order to win hands, what do you have to do? In order to win hands a lot, it would normally be the same wads when it stays in the hands all the time. But in order to win, win at a higher percentage, it means you have to stay in the game longer. Really, really good poker players know one thing. Win to get out. Think about that with your software. Do you know when to get out? Are you playing a losing hand all the way to the end? Something to think about. Are you looking for the right metrics? I mean, I compare this one often to on-time delivery. Does on-time delivery really matter? Or does it really matter how much value you deliver? What's your focus? Oil and gas exploration. Something that's near and dear to my heart. What do you think the success rate is for new frontier oil and gas exploration? High success? Any ideas? Less than 1%, 2%, 10%. Yeah, so it's about 90% to 80% failure, 10% to 20% success rate. 90% failure if you're using competitors' software and data. 80% if you're using IHS data. We make a big difference and products we produce. Oil and gas also has a lot of safety issues. It was my beginning to think it wasn't such a good idea to turn off those unit tests. You've all heard about the big blowout that BP had in the Gulf of Mexico and certainly we heard about it and I think probably everyone heard around the world. It was huge. Actually, the company I was with, Halliburton, was involved with it. I'm not speaking from any inside information but basically, it was effectively a decision by the engineers to turn off some unit tests that resulted in the big explosion. They did some tests. They didn't quite get the results they wanted. They decided they weren't going to run the second test and they weren't going to run the follow-up test that would have actually caught the issue and they would have known about it and they would have been able to shut down so that they didn't have the explosion. It's in software, it's an issue, sometimes everywhere we go. It's great and lintening to those tests is important. One of the things we do in Welling Gas is use something called a tornado plot. In other industries you see this. I'm going to introduce this idea of a tornado plot and what you're really looking at here is a plot that shows what are the uncertainties that are really important. This is net present value that we're showing and the base case is the vertical line there. It's 1,350 million is the base case and then what we plot is this tornado that shows where is the uncertainty. So the big uncertainty in this is in the amount of reserves in the ground which could take our valuation way down to possibly only 300 million or it could take us up into a much larger amount. Next down we have Welling Gas prices and then there's uncertainty and then down at the final at the bottom there is drilling cost. Now this is one of those things where we often have the measurement inversion. What do we do? Well we focus on cost sometimes. It's not that important. What really matters is what's the big picture. So what would an intelligent company do with this? They say well this is my size and my uncertainty. What could I do to understand my uncertainty better? So like with reserves maybe I could get a little bit more data about it. Maybe I could do some more seismics or maybe I could do some preliminary drilling to find out and quantify a little bit better what reserves are so that I can better understand the amount of uncertainty and make informed decisions based on that. So now I'm going to go into some risk management tools associated with that in particular the one particular flavor of this that I like and I've worked with Chris Madsen Olaf Masson on real options. And the idea behind real options is that it's the right but not the obligation to undertake a certain action prior to an expiry date. Very simple concept. You have the right to do something but you don't actually have to do it and at some point that right will expire and you no longer can do it. One very simple example of a real option is an airline ticket. I buy an airline ticket doesn't actually don't even think about whether it's a refundable ticket or a nonrefundable ticket because that gets into a secondary issue. Let's just say I'm going to buy a ticket and given I've got that ticket I now have a right to board the plane. They can't take that away from me because I've got that right to board the plane. Now there's other issues associated with if they're not going to fly and things like that but basically I have a right to board the plane. They can't make me get on that plane. If I don't want to get on the plane I don't have to. So I've got a right to do it but I don't have to do it and at some point the plane is going to fly away and my option to get on that plane has expired. Very simple, simple concept. Most tickets, almost all tickets operate as a form of a real option and if you want to look at it from the perspective of a refundable, nonrefundable ticket what you actually have is an option on an option. You have the option to get for the ticket you also have the option to refund it and so you have the option on the option. Now there are a couple of tickets that aren't real options. Anyone have an idea of a ticket that's not a real option? What's that? Yeah? Speeding ticket? Parking ticket? Not necessarily options. Alright. So real options are the right but not the obligation to take some action prior to an expiry date. Chris and Olaf have simplified this into three bite-sized pieces that I'm going to expose here. First of all, options have value. There's a value to having that option. Options will expire at some point and key one, you never commit early unless you know why. So it's all about there's a decision you can make or you can commit and the commitment actually expires the option. So you don't want to do that unless you know why. But if you know why, you may very well want to commit early. I'm going to take a different spin on this but first we're going to look at options from the perspective of options, decisions, and commitments. So options are the options that are available to you. From that you can make decisions. Decisions don't necessarily expire the option. You can make decisions but the decision is generally the difference between decision and commitment is decisions are reversible. So you still have the option it's just you happen to go down a particular decision path and you actually want to be able to do that. What you want to make sure you don't do is commit early unless you know why. Because once you make the commitment you've expired the option the value of that option has disappeared so you've made that commitment. The commitment might be the right answer but you want to know why you've made that commitment if you're going to do it. So we can look at this as an example here someone's going to run and jump off the cliff. At the beginning they have an option. They have the option of jumping off the cliff or not. Then they've made a decision, a decision to start forward. At some point they've committed and in this case it's probably committed pretty early. Okay? But this is the idea of the commitment of which point you can't go back. So given that framework of real options there's four particular areas that I want to call out and put it in different context. First of all is the value of uncertainty. Second value of information. Third is the value of flexibility and the fourth is the cost of delay. So the first point is the value of uncertainty. This guy Patrick Leach an orthopathic leach why can't you just give me a number? I love that title. Why can't you just give me a number? Because it's something we face all the time. As a very provocative question what's the source of all value? He teaches a lot. He's asked of all of his classes. He never comes around to people giving him the answer that he's looking for. It's not really surprising that he doesn't get it. He gets the standard answers to all of the questions and all of those things which are very, very truly do contribute to value. What he's really trying to get to what's the source of all value though is the fact uncertainty has to be present for value to be created. There has to be uncertainty because if there's not uncertainty if you have a competitive advantage and you're doing something and you're building value it'll be arbitraged away and eventually it'll be risk-free. So the fact that there's uncertainty is the source of value. And this is actually shown out also in Black Scholes. Black Scholes turns out to be not a very good way of looking at real options. It turns out to be a pretty good way for pricing financial options. But one of the key things from it is there's a volatility number a V number, a volatility. Volatility is an indication of how great the uncertainty is and the uncertainty the higher the price. So the higher the value. The amount of uncertainty has a direct input indicator of the amount of value. This is also shown here as a risk-reward, risk-return chart here. And typically this is the best you can do. That efficient frontier is the best you can do. At a high risk you're going to get a higher return. You can have something under the efficient frontier which means you've got a high risk and not getting as much return as you possibly could be getting. But you generally can't go to the left of that curve. The thing I want to pull out of here is the value of certainty is that if you're going to reduce the risk you're typically also going to reduce the return. So I'm going to go through an example here using a tornado example and just saying we're looking at this and we've seen this chart here and we say so our example here product acceptance uncertainty how well we're going to do in the market that's our huge uncertainty. We've got some schedule uncertainty we've got some general market uncertainty and maybe we've got some cost uncertainty. And we look at that and there's some negative numbers there and we say you know what we'd really like to do some things to reduce our uncertainty and we're going to eliminate those red bars okay so very simply management okay let's remove those red bars so we no longer have low end well what's really happening in the system we can probably do that but the result is going to be we're going to chop away all the high end so yes we can reduce uncertainty but reducing uncertainty has actually reduced