BEWARE: The Backlog Trap - Avoid Common Challenges When Picking Construction Projects

(SPEECH)
But our infrastructure study-- we have 224 projects in total. We used cut-off of contracts over $250 million in contract amount. So the average contract value of these 224 projects is $800 million-- you're going to say, wow, those are really big, I never do those and that's fine. Again, this is a study. We're just trying to learn from it, and then we're going to apply it from your business.

The types of projects we looked at were bridge, highway, rail/light rail, tunnel, and heavy other. And we looked at it by procurement type, and contract value, and type of work, overall project risk, and design risk. Those are factors that we looked at in trying to figure out what it would tell us. We really didn't know. We had some ideas but we weren't sure. We were going to let the data speak for itself.

(DESCRIPTION)
Image, A line graph with three solid lines and one dotted line

(SPEECH)
So this slide is kind of busy so let's start in level set here with this blue line. So this blue line is the initial weighted average-- a gross margin that we receive on WIP reports from our contractors, OK? And you'll see these contract years are here. So if we use 2009, for example, any project started in 2009 shows up here. Any project that shows up in 2015 shows up here. We don't move them. This is by contract year. So compared to the blue line, this is what they thought they were going to make when they did it, or at least put it on their bid schedule.

There is a black line here that separates from the red dotted line and that is what we call the latest margin. So as you can see back in here, these jobs are all done and they match. But in a few cases, there is separation between the black and red lines. And those are what we call-- we subtract out under-billings. And for those of you familiar with construction, under-billings, typically, are code named for claims that aren't fully realized. So we actually think the final margin will sit somewhere between the black and the red lines for these and other works.

The other thing to think about is these jobs right here are not fully baked. So we really don't know how they're going to perform. We're optimistic that that's going to be a more positive trend. But clearly, from 2018 till 2011, these results were not very good, right? And you can see the bottom peaks of average losses. Like 5% and 4%, that's below your cost and before overhead. So not a very good result. If you really think about getting a profit, I would draw a line at 5% here and that's kind of an average overhead number, and you would need to be above that to actually make money if you did this work. So kind of an interesting slide. So clearly, that seven-year period I showed you impacted the results.

(DESCRIPTION)
Text, Margins by Contract Value

(SPEECH)
Now we looked at it by contract size, OK? And if you looked at that-- again, the same lines-- blue, black, and red you're familiar, and we separated them by average contract size, all right? And you can see that this is the 550 to a billion, this is over a billion. And you can see there's a big drop down once they hit 550, right? They went from the 8% to 9% range to the 5% to 6% of range for 550 to a billion. But then when they got to a billion they actually creeped up a little bit though there is some claims there. But clearly, the contractors were not making at any level what they thought they were making at bid time. They just weren't performing as well. But obviously, these bigger projects performed clearly but they weren't clearly correlated because if they were, this line would have kept going down. So it isn't just contract size that drives it, though it clearly seems to have some impact once you get to this level.

(DESCRIPTION)
Graphs, Margins by Types of Work

(SPEECH)
So then we looked at types of work. So the pie charts on the left break out the works by number of project. The largest class is highway, followed by rail and light rail here in the light green, and then bridges came in third. Here in the middle, we had tunnel work-- this is stand-alone tunnel work, and finally heavy other. And heavy other would be things like dams, and runways, and that spillway I showed you, and unique jobs, levees and things like that.

But when you looked at them by contract value, you can see that the rail and light rail get bigger and everyone else kind of holds its own other than what they give back to the rail and light rail, and obviously, highways give the most. So these are the biggest contracts that are out there. These infrastructure projects where they're building new rail like in California or the Purple Line in Maryland. There's many of them in many major cities.

So and then on the right, you can see the average contract size for each type. And again, rail and light rail far and away the biggest for that. But all of these are big classes of business.

So now let's look at margins, OK? The dark blue bar-- that's the original margin. The gray or light blue-- however you see that-- that's the final margin on average, again weighted. And what's interesting-- I'll draw your attention to is what areas perform the worst? Well, highway jumps out at you, right? And everyone would tell you highway seems like probably the easiest of all these types of work, but they had the lowest initial margin, the lowest final margin. And the same for bridge. Both of these fell off around 5% from their initial bid. And neither of them made a lot of money, which was very surprising. Remember this is a 17-year study-- 224 jobs in all.

Rail and light rail fell off about 30%. And the areas that performed the best were tunnels surprisingly. That really surprised us because that is considered one of the riskiest classes of business, but that only fell off a point. That obviously would bid it at a much higher expected margin. And then heavy other-- that kind of catch all category. That also fell out a point. A lot of that work-- believe it or not-- was Katrina work in New Orleans when they redid the levees and most of that work did pretty well as evidenced by here. But these three here did not do well. But again, you couldn't just say, hey, it's just the type of work thing. Maybe unless it's anything highway related.

(DESCRIPTION)
Graphs, Procurement Type

(SPEECH)
So then we looked at procurement type. And if you look at procurement type, because these are big jobs 44% were design build and another 15% were P3. So that's about 60% all together. Then 33% is design bid build and that's the old riff and read-- that's what most people are used to. The owner does the design, the contractor reviews the design, bids the work, and builds it, right? That's that type of work. And that's about one in three jobs.

And finally, even in the highway business, CM/GC, which is very common today in the building world has started to get a little bit of a toehold in that world, and we saw a little bit of that, but it's only 7% of what's going on and only 5% of the contract value. And you can, again, see how much of the contract value the design build makes when you throw in the P3-- that's 65 to 75. That's almost 75% of all jobs fall to this. It's like 74 and 26. So clearly, the design build is the procurement type for these big jobs.

So as you look at contract sizes as you can see the design build and CM/GC-- they were the smaller jobs on average and the design build and the P3-- those are the bigger jobs on average. And you can see those private contract prices. So how did this perform?

(DESCRIPTION)
Graphs, Margins by Procurement Type

(SPEECH)

Well, this is where it really gets interesting. So let's start with the CM/CG. Again, initial margin in the royal blue, the final margin in the gray. And you can see a 5% improvement. We hadn't seen an improvement yet when we started this slide, had we? Or this presentation. So first area these jobs weren't really well. Design bid build only dropped about a point. That really surprised us. Everyone said, oh, design bid build is a mess, it can't be done. But clearly, it performed well. Design build oh, this is the best one, everyone does great at that. Well, not really. Again, a 40% drop off from the average initial margin to 6%.

And lo and behold, when you went to P3 what did you see? Over a 6% drop, it dropped 2/3. The margin from what you bid to that. And if you throw in that initial 5% of overhead, P3 work in general and the US for infrastructure work has lost money. Lost quite a bit of money for the contractors. It's not been a good procurement method.

(DESCRIPTION)
Charts, The Impacts of Design Risk

(SPEECH)
So to take another slice of that, we looked at what about jobs where contractors had design risk and jobs where they didn’t. And on the left, you can see the average margin for each type of job in the infrastructure space, where they had risk and where they didn’t and there's a 6% difference almost-- a little over 6% difference.

And then finally, we look, OK, when they did that, how much did their margin change? Well, for no design risk margin actually improved by 12% from the initial bid, but with design risk margin dropped over 76% on average per job. So clearly, the issue is procurement type, design, build, and infrastructure. And the question is, why is that the case?

(DESCRIPTION)
Graph, two solid lines and one dotted line

So that's where the correlation is. So the answer to our question was actually procurement type. Probably, job size was secondary. And all of these jobs are long, so duration was kind of a trick question. But yeah, the biggest driver of it is procurement type.

We show it to you by risk. Least risky jobs over here-- CM/GC then CMAR-- kind of the same thing-- then design bid build right in the middle, design build and then P3. And you can kind of see, the ski slope-- it starts here once you hit design bid build, and it really drives them off and down.

So a little bit of surprise to us. And the question we ask is why do you think that was the case? And I really think it has to do-- because you probably have some building contractors out there that say, I do well with design build. But in the infrastructure world, especially with P3, they're asking the contractors to price the design risk and the whole contract, including all quantities on about a 30% design. And what our data clearly shows us is contractors can't do that. It's just impossible to do. And they're losing lots of money and that's why I remember all those statements I read at the beginning of the thing. That's the big driver-- is they were just pricing work too early where they couldn't do it.

(DESCRIPTION)
Charts, Profitability by Procurement Type

(SPEECH)
And here's an example. So this is the profitability by procurement type. So again, red would be a loss, blue 0-5 yellow 6-10, green 10 plus. And really, 5 is kind of a loser job in these things when you put overhead on it. So you can see in CM/GC every job made money. And 2/3 made pretty good money. And the other thing there is no dispute.

Now if you look at design bid build, about 75% of the jobs made money and about 25% struggled with that. Not perfect but maybe that's palatable-- 3 out of 4 were pretty good jobs. But you get into design build 42% of the jobs-- 42% didn't make their adequate margin of more than 5%. And a full 19% lost money-- that's double what design bid build.

And then if you get into P3, it gets even worse. Almost 50% of the jobs-- 1 in 2 didn't make an adequate return and 34% just flat out lost money. 1 in 4 made an acceptable return, versus 2 out of 3 in CM/GC. Not exactly what, I think, a lot of contractors thought would have happened when they started pursuing this work.

(SPEECH)
So as we wrap up this infrastructure study and move it to you guys-- a couple of things to remember. Procurement type does matter, and when you get to your cost-- does matter. The construction industry clearly has not been able to price risk and design build and P3-- and we believe a big factor is that they're being asked the price of one design is roughly 30% complete. If you're bidding something with higher risk, you've got to increase your pricing. It isn't going to work any other way. And don't forget that the preponderance-- that's a tough word to me today, sorry about that-- that should be a concern to both owners and contractors. They should want more predictable outcome.

And finally, CM/GC-- or what we call the delayed GMP or gross maximum price or guaranteed maximum price-- for progressive design build model, where you get at your pricing later in the procurement process provide better, more consistent results for both owners and contractors. So that's our takeaway from this, and we're going to be actually advocating for that in the marketplace, especially in light of the infrastructure study coming out.

(SPEECH)
And what we call risk creep is undetected increase in the level of risk in the backlog that occurs over a period of time. So let's compare it with something. We have a few examples here that you can see. The hour hand on the watch, assuming you have a traditional watch. If you stare at it, it doesn't seem to move. But if you stare at it for 60 minutes, it goes all the way around in a complete circle. But we just don't tend to notice that.

Or take your kids, for example, maybe you have that place on the wall where you measure them, and they don't seem to grow on a day to day basis but when you measure them each month, you can see them getting taller, and then they become teenagers, and then they go to college-- and in the case of my kids, now they're out in their own homes and have jobs, and it happens really fast.

And then maybe the best example is the Navy SEALs here on the right in the ghilly suits, that's what you're seeing there. And these guys are fully aware-- if they move slowly enough and don't draw attention to themselves that they can get to the spot they need to get to and do what they're set out to do. In this case, it would be very good for you if they were looking for you.

So these are all examples of risk creep. And that's what can happen in your backlog. And that's what we think happened to these infrastructure contractors. We don't think they knowingly lost money but that happened to them.

(DESCRIPTION)
Text, Risk Creep Illustrated

(SPEECH)
So let's think about how that can happen in an illustration here for that. So let's start with this slide. This actually the slide with how we price work. Our price-- our business in surety, OK? There is a market price. Everyone knows that. That's out there, OK? And there is a risk-based price that's out there somewhere-- a theoretical risk-based price and where they cross, right? Sets up a wall. So you certainly want your risk-based price to be lower than the market price because that's how you make your money, right? And so ideally, you want to kind of operate over here. That doesn't mean you can't ever make money if you're operating out here, but it's harder to do that because the things that set it off may not allow you to do that.

So let's look at a job. The single dot that we have right here that you just acquired, and you can see it's just inside where everything works for you, OK? It's a good job, It's right out where the price point to the two curves pass, and it's a good job. Let's say what happens is that-- and this happened in the infrastructure study. They had a recession and a bunch of contractors started deciding to chase bigger jobs, and then some contractors in Europe where things were worse than the US decided to come over here and pursue work and all of a sudden the risk curve moved. But you didn't change but the market changed, right? And now there's more competition. So the market price slid down.

(DESCRIPTION)
Image, The dot shifts downward.

(SPEECH)
Now the owner starts to see more competition-- what's the first thing they want to do? Or shift more risk to them. So now that risk-based curve is going to move up and even though your dot hasn't moved, it's further out the curve because the red line and the green line now crossed way back here. So all of a sudden this one job you had that was a good job, now might not be such a good job anymore unless you adjust your pricing.

(SPEECH)
A couple of key points here. Hey, if the market's not great, it's OK to get smaller. Avoid losing jobs-- I would tell you to do that. And don't venture too far from the bull's eye, right? Those are three things I would take away from this slide as we move forward. I want to have some time for Q&A. And then finally, don't forget the market price for risk will go down, but it doesn't make it the right price. The market's not that smart. That's really important to remember.

Another good thing in bad times is a bad owner equals a bad job almost every time. And even though you really need to work, those odds don't change. And if you are going to do something new there is a learning curve, and you should price that in or be willing to invest that money.

(DESCRIPTION)
Text, Bad Projects

(SPEECH)
So let's think about how we look at work and how we train our underwriters to look at work. So this is a really simple example, but something for you to look at. So we know by our history that bad jobs have some common theme and most of that is they have too many new things, which we're going to call the News. For example, a new job size, and geography, a new owner, a new end market that they've never worked in, new procurement methods, new risk/return expectations. All the things those are News. And if you add a point for each one, you can build some type of score.

(DESCRIPTION)
Text, Staying on Target

(SPEECH)
So this is our project A. Normal work for a new customer in a new geographic market. Well, clearly there are two News here, so boom, that one got 2.

Now this project B, normal work for a great customer in a new geographic market. Well, that would be 1, right? And then finally, this is a fantasy project. But this is a bigger project than this contractor has ever done before. It's a P3. They also haven't done it before. They never worked in Miami, and it's their first bridge. So that's obviously 4 News and obviously anyone would think that's a good job-- hasn't been listening to my report just yet or this presentation.