Economics of shipyard

[atpollard]

This may be 100% not what you are looking for, but I approach these sorts of problems from the opposite direction using a crude parametric from the real world.

So some typical values are ...
parts = 50% of final cost
labor = 30% of final cost
plant/facility = 10% of final cost
profit = 10% of final cost
(these figures vary with industry and product, but these are good, conservative, middle of the road values)

So lets say my Starport wants to build MCr 10 of starships per day.
I need to import MCr 5 of finished parts per day to build them from.
I need to pay MCr 3 per day in salaries for workers.
I need to pay MCr 1 per day to maintain the physical structure to build the starships.
I will earn MCr 1 per day in profit (which may be shared with stockholders if I had to borrow money to build all of this).

So let's scale those costs up to monthly (28 days):
Revenue = MCr +280 worth of starships
Parts = MCr -140
Labor = MCr -84
Plant = MCr -28
Profit = MCr 28

Let's assume that I want to put some meat on those bones. Start with some simple assumption for workers, I like 5 at skill-0 (Cr 1000/mo) for each 1 at skill-1 (Cr 2000/mo). Then 5 at skill-1for each 1 at skill-2 (Cr 3000/mo). Then 5 at skill-2 for each 1 at skill-3 (Cr 4000/mo). So a typical 'team' is 1 @ skill-3 (Cr 4000/mo), 5 @ skill-2 (Cr 15k/mo), 25 @ skill-1 (Cr 50k/mo), 125 @ skill-0 (Cr 125k/mo) = Team of 156 workers @ MCr 0.194/month. So my MCr 84 of monthly labor requires about 433 Teams or 67,548 total workers.

The Physical Plant costs MCr 28 per month, and I like to use the Starship Finance rules to estimate facility costs. So start with 1/240 cost per month and multiply the MCr 28 plant cost per month x 240 to arrive at MCr 6720 worth of Shipyard. I assume that the yard is never paid off because after 40 years, the MCr 28 per month just goes to replacing and overhauling systems that have worn out.

Since you wanted to do everything, use the PARTS from the Shipyard as a starting point for a factory to build the parts:
Revenue = MCr +140/mo worth of starship parts
Processed Materials = 50% of final cost = MCr -70/mo
labor = 30% of final cost = MCr -42/mo = 216.5 Teams = 33,774 workers
plant/facility = 10% of final cost = MCr -14/mo = MCr 3360 factory
profit = 10% of final cost = MCr 14/mo

And use the PROCESSED MATERIALS from the Factory as a starting point for a refinery operation:
Revenue = MCr +70/mo worth of processed materials
raw ore = 50% of final cost = MCr -35/mo
labor = 30% of final cost = MCr -21/mo = 108 Teams = 16,848 workers
plant/facility = 10% of final cost = MCr -7/mo = MCr 1680 refinery
profit = 10% of final cost = MCr 7/mo

And use the RAW ORE from the Refinery as a starting point for a mining operation:
Revenue = MCr +35/mo worth of raw ore
supplies = 10% of final cost = MCr -3.5/mo
labor = 30% of final cost = MCr -10.5/mo = 54 Teams = 8,424 workers
plant/facility = 50% of final cost = MCr -17.5/mo = MCr 4200 mining equipment
profit = 10% of final cost = MCr 3.5/mo
[Mining gets the 'parts' mostly for free, it is heavily dependent on specialized heavy equipment ... so the percentages for parts and plant are reversed.]

Obviously, MCr 10 per day is a larger operation than you had intended, but the concept works with other values. Just scale it down a couple orders of magnitude. If you have TL impact average wages, then TL will reduce the number of employees as technology makes each more expensive and productive.

Anyway, this is just another idea to use or ignore as seems best to you.

 

[steve98052]

This may be 100% not what you are looking for, but I approach these sorts of problems from the opposite direction using a crude parametric from the real world.

So some typical values are ...
parts = 50% of final cost
labor = 30% of final cost
plant/facility = 10% of final cost
profit = 10% of final cost
(these figures vary with industry and product, but these are good, conservative, middle of the road values)

I'm not sure how those figures align with real manufacturing -- I could probably download a few annual reports to check -- but this is definitely a better approach to the problem if any sort of realism is intended.

Note that automation increases a lot as technology advances. US manufacturing employment peaked around 1970, and except for the past six years it has declined almost every year since that peak. But in spite of that employment decline trend, production continues to increase (except during deeper recessions). Production per employee today is about tripled since (must look up date again), but total production has doubled in the same time period, so employment is about two-thirds of what it was.

Stretch this across Traveller tech levels, which the trend probably would, to the point that manufacturing workers in a medium stellar world would likely be as scarce as farmers in today's economy, even though farmers were once 90% of the population.

Obviously, MCr 10 per day is a larger operation than you had intended, but the concept works with other values. Just scale it down a couple orders of magnitude. If you have TL impact average wages, then TL will reduce the number of employees as technology makes each more expensive and productive.

The MCr 10 example might actually be uncompetitive, because of labor costs. If the trend for manufacturing employment follows a plausible trend, a factory of high enough tech level to be workable in a space station might have only one manufacturing employee in each of just a few job categories, capable of running a huge factory, so having one employee in each job category in a small factory might run up labor costs too much. (On the other hand, it may not be a big problem to run 2% factory labor instead of 1%.) The overwhelming majority of employment at higher technology levels will be marketing, sales, lobbyists, engineering, and so forth -- not factory workers.

 

[atpollard]

This may be 100% not what you are looking for, but I approach these sorts of problems from the opposite direction using a crude parametric from the real world.

So some typical values are ...
parts = 50% of final cost
labor = 30% of final cost
plant/facility = 10% of final cost
profit = 10% of final cost
(these figures vary with industry and product, but these are good, conservative, middle of the road values)

I'm not sure how those figures align with real manufacturing -- I could probably download a few annual reports to check -- but this is definitely a better approach to the problem if any sort of realism is intended.

Note that automation increases a lot as technology advances. US manufacturing employment peaked around 1970, and except for the past six years it has declined almost every year since that peak. But in spite of that employment decline trend, production continues to increase (except during deeper recessions). Production per employee today is about tripled since (must look up date again), but total production has doubled in the same time period, so employment is about two-thirds of what it was.

Stretch this across Traveller tech levels, which the trend probably would, to the point that manufacturing workers in a medium stellar world would likely be as scarce as farmers in today's economy, even though farmers were once 90% of the population.

Obviously, MCr 10 per day is a larger operation than you had intended, but the concept works with other values. Just scale it down a couple orders of magnitude. If you have TL impact average wages, then TL will reduce the number of employees as technology makes each more expensive and productive.

The MCr 10 example might actually be uncompetitive, because of labor costs. If the trend for manufacturing employment follows a plausible trend, a factory of high enough tech level to be workable in a space station might have only one manufacturing employee in each of just a few job categories, capable of running a huge factory, so having one employee in each job category in a small factory might run up labor costs too much. (On the other hand, it may not be a big problem to run 2% factory labor instead of 1%.) The overwhelming majority of employment at higher technology levels will be marketing, sales, lobbyists, engineering, and so forth -- not factory workers.

Automation is the counterbalance for labor, but will just shift the percentages from Labor to Plant costs. These Labor figures were for more labor intensive operations like a TL 5 factory, or a modern repair shop or custom bike shop. I think that a 10% profit margin is in the right ballpark. I think GM operates on a 4% margin. So for a space station, the Cr 2000 per person Life Support will make maximum automation an easy decision, so perhaps 10% labor and 30% factory might be better with a typical salary in the Cr 5000 per month range for the sort of high tech specialists that run a robotic manufacturing operation.

Another way to model that level of automation is just to pay Cr 100,000 per worker for robot workers and finance them at 1/240 cost per month. Then you just need to figure out how many technicians you need to maintain all that automation and technology. Starship Engineer requirements give a starting point for guesstimates.

 

[AnotherDilbert]

So some typical values are ...
parts = 50% of final cost
labor = 30% of final cost
plant/facility = 10% of final cost
profit = 10% of final cost
(these figures vary with industry and product, but these are good, conservative, middle of the road values)

This might be true for the factory, but probably not for the entire company, and hence total cost of the product.

Customers will not spontaneously line up outside the factory to cart away the produce as it is finished, and pay cash immediately. You need a lot more than just a factory in order to sell vehicles.

Remember the friendly General Products and GSbAG dealers in a nearby space station will do their best to take your customers and drive you out of business. They will offer competing small craft with extra chrome stripes designed by a famous artist on easy instalment plans. They can also spread the R&D costs on millions of units Imperium-wide.

Government will also be highly interested in your operation, not just taxes, but also labour laws, 'elf&safety, fire control, environmental, food preparation and handling, forbidden technology, end-user certificates, and money-laundering laws (you can prove that the customer's money was legit and taxed, right?). Some of the inspectors might want bribes. What happens when your product is used by a pirate to kill a nobleman's cousin?

Labour unions might blockade the company, unless you raise salaries, pensions, and hire someone's nephew.

Criminals will try to steal your equipment, product, or perhaps your entire space station. Or they want protection money?

Anti-war or robot emancipation campaigners might picket or sabotage your operation.

All of this have to be handled. By you.

 

[atpollard]

So some typical values are ...
parts = 50% of final cost
labor = 30% of final cost
plant/facility = 10% of final cost
profit = 10% of final cost
(these figures vary with industry and product, but these are good, conservative, middle of the road values)

This might be true for the factory, but probably not for the entire company, and hence total cost of the product.

Customers will not spontaneously line up outside the factory to cart away the produce as it is finished, and pay cash immediately. You need a lot more than just a factory in order to sell vehicles.

Remember the friendly General Products and GSbAG dealers in a nearby space station will do their best to take your customers and drive you out of business. They will offer competing small craft with extra chrome stripes designed by a famous artist on easy instalment plans. They can also spread the R&D costs on millions of units Imperium-wide.

Government will also be highly interested in your operation, not just taxes, but also labour laws, 'elf&safety, fire control, environmental, food preparation and handling, forbidden technology, end-user certificates, and money-laundering laws (you can prove that the customer's money was legit and taxed, right?). Some of the inspectors might want bribes. What happens when your product is used by a pirate to kill a nobleman's cousin?

Labour unions might blockade the company, unless you raise salaries, pensions, and hire someone's nephew.

Criminals will try to steal your equipment, product, or perhaps your entire space station. Or they want protection money?

Anti-war or robot emancipation campaigners might picket or sabotage your operation.

All of this have to be handled. By you.

And does any of this not apply to the license to print money presented in the OP?
A small manufacturing module, a few workers, some mining bots and a huge gap between revenue and costs.

Perhaps that is why LABOR is 30% of REVENUE ... all of the Corporate Lawyers and Accountants and Marketers that deal with those problems on a corporate scale.

 

[Condottiere]

I suspect it depends on how each system government regulates asteroid mining and prospecting; it might be a monopoly, you might need a license, you get to pay a hefty tax to the Exchequer.

Manufacturing spaceship components might require some kind of supervision to ensure the components follow engineering standards and official regulations.

Gremlins might build a lair in your factory.

 

[AnotherDilbert]

And does any of this not apply to the license to print money presented in the OP?

Yes, certainly it does.

A small manufacturing module, a few workers, some mining bots and a huge gap between revenue and costs.

The OP assumes you sell everything you produce immediately, from the factory floor at retail price with no additional costs.

I'm trying to say:
Factories do not sell at retail price.
Factories do not automatically sell everything they produce.
Customers do not pick up products from the factory floor the instant they are finished.
You do have additional costs.

Hence factories are not licenses to print money.

 

[atpollard]

And does any of this not apply to the license to print money presented in the OP?

Yes, certainly it does.

A small manufacturing module, a few workers, some mining bots and a huge gap between revenue and costs.

The OP assumes you sell everything you produce immediately, from the factory floor at retail price with no additional costs.

I'm trying to say:
Factories do not sell at retail price.
Factories do not automatically sell everything they produce.
Customers do not pick up products from the factory floor the instant they are finished.
You do have additional costs.

Hence factories are not licenses to print money.

Oops. Sorry, I didn't see it that way. My bad.

Starships are a strange thing since they are all custom made, like a Learjet or a new house. There are no middle men because each one is made to order and custom built. You are buying from the factory (so to speak).

In the real world, there are massive savings from mass production. Something on the order of each time you double the number of identical units produced, the per unit cost drops by 5% (95% of current cost). So throwing some numbers in there, if the first starship cost MCr 100, then ...
ship number 2-3 costs 100 x 0.95 = MCr 95 each
ship number 4-7 costs 95 x 0.95 = MCr 90.3
ship number 8-15 costs 90.3 x 0.95 = MCr 85.7
ship number 16-31 costs 85.7 x 0.95 = MCr 81.5
ship number 32-63 costs = MCr 77.4
ship number 64-127 costs = MCr 73.5
ship number 128-255 costs = MCr 69.8
ship number 256-511 costs = MCr 66.3
ship number 512-1023 costs = MCr 63.0
ship number 1024-2047 costs = MCr 59.9
ship number 2048-4097 costs = MCr 56.9
ship number 4096-8191 costs = MCr 54.0
ship number 8192-16383 costs = MCr 51.3

So a big factory will bankrupt a little factory by selling more units for less and driving down the price.
But who places an order with a shipyard for 100 absolutely identical far traders?

Now a factory selling 10,000 identical Ship's Boats per year at 52% of the single unit cost wants to sell them in batches of 100 to 1000 units to wholesalers that will resell them at 75% of the single unit cost in units of 5 or more to local dealers who will sell you just one for 90% of the single unit cost. You would need to order 8 or more from the local shipyard to get a better price. So the local shipyard doesn't bother making Ship's Boats because there is no demand.


If you wanted to use my figures to model your mass-production and distribution, just add 2 more levels above the factory:

Dealer sells MCr 1 Air rafts:
-500,000 wholsale price of air raft
-300,000 Salaries of employees at dealership (sales, service, financing, government paperwork, marketing, accounting, security)
-100,000 Dealership Building
+100,000 Profit

Wholesaler sells MCr 0.5 Air rafts:
-250,000 factory price of air raft
-150,000 Salaries of employees (warehouse, security, accounting, sales, transport)
-50,000 Warehouse Building & Distribution Transports
+50,000 Profit
(this is per air raft sold, so if the Wholesaler supplies 10 Dealers, then multiply by 10)

Factory sells MCr 0.25 Air rafts:
-125,000 parts
-75,000 Salaries of employees at factory
-25,000 Factory
+25,000 Profit
(this is per air raft sold, so if the Factory supplies 100 Dealers, then multiply by 100)

 

[AnotherDilbert]

So a big factory will bankrupt a little factory by selling more units for less and driving down the price.
But who places an order with a shipyard for 100 absolutely identical far traders?

Quite, but if enough Far Traders are sold each year the Megacorp ship builders will have a production-line per sector or so producing a steady stream of them, perhaps undercutting local yards.

If you wanted to use my figures to model your mass-production and distribution, just add 2 more levels above the factory:

From the number I have seen margins vary wildly, e.g. 50%-60% for clothes down to 10%-15% for consumer electronics.

Just assuming a 50% margin is perhaps a good first order approximation, but I suspect ships will have a lower margin.

 

[phavoc]

Big ticket items dont have the same sort of margins as a tshirt or printer cable. The more costly the item or service the lower the margin, generally speaking.

You also have to look at volume. An out of the way planet with a class A shipyard isn't going to have the volume that day a yard on Regina or some other hub planet with a large population. Volume helps reduce costs because you spread fixed costs across more ships.

It is a futile effort to try and I insert all the nuances of a real economy into Traveller, which decidedly does NOT have a good economic model.

 

[atpollard]

Big ticket items dont have the same sort of margins as a tshirt or printer cable. The more costly the item or service the lower the margin, generally speaking.

You also have to look at volume. An out of the way planet with a class A shipyard isn't going to have the volume that day a yard on Regina or some other hub planet with a large population. Volume helps reduce costs because you spread fixed costs across more ships.

It is a futile effort to try and I insert all the nuances of a real economy into Traveller, which decidedly does NOT have a good economic model.

Agreed.
It also misses the point that Traveller is a game, so the most important thing is not accuracy or detail, but having fun. Fun will depend on what the Ref and players want out of their investment in time. For some, detail will be important. For others, it will matter not one iota.

I shared the level of detail I use when I need detail.
Good luck finding a level of detail that you are comfortable with.

 

[Tom Kalbfus]

And does any of this not apply to the license to print money presented in the OP?

Yes, certainly it does.

A small manufacturing module, a few workers, some mining bots and a huge gap between revenue and costs.

The OP assumes you sell everything you produce immediately, from the factory floor at retail price with no additional costs.

I'm trying to say:
Factories do not sell at retail price.
Factories do not automatically sell everything they produce.
Customers do not pick up products from the factory floor the instant they are finished.
You do have additional costs.

Hence factories are not licenses to print money.

Oops. Sorry, I didn't see it that way. My bad.

Starships are a strange thing since they are all custom made, like a Learjet or a new house. There are no middle men because each one is made to order and custom built. You are buying from the factory (so to speak).

In the real world, there are massive savings from mass production. Something on the order of each time you double the number of identical units produced, the per unit cost drops by 5% (95% of current cost). So throwing some numbers in there, if the first starship cost MCr 100, then ...
ship number 2-3 costs 100 x 0.95 = MCr 95 each
ship number 4-7 costs 95 x 0.95 = MCr 90.3
ship number 8-15 costs 90.3 x 0.95 = MCr 85.7
ship number 16-31 costs 85.7 x 0.95 = MCr 81.5
ship number 32-63 costs = MCr 77.4
ship number 64-127 costs = MCr 73.5
ship number 128-255 costs = MCr 69.8
ship number 256-511 costs = MCr 66.3
ship number 512-1023 costs = MCr 63.0
ship number 1024-2047 costs = MCr 59.9
ship number 2048-4097 costs = MCr 56.9
ship number 4096-8191 costs = MCr 54.0
ship number 8192-16383 costs = MCr 51.3

So a big factory will bankrupt a little factory by selling more units for less and driving down the price.
But who places an order with a shipyard for 100 absolutely identical far traders?

A Far Trader is a standard design, therefore it is built in a factory to reduce cost, the customer buys one and then customizes it later at additional cost to himself. The closest equivalent to a starship factory would be an air plane factory, the manufacturer has an assembly line and sells the basic model, another department in that same company or a different company customizes them. Typically customization is done in a shipyard, but the basic model, if its a standard model is built in a factory, and there is an assembly line which produces these starships, they are shipped to various shipyards for customization. A shipyard can also produce a nonstandard custom starship or spaceship.

 

[phavoc]

Something that complicates the model (gee, ANOTHER thing??) is that the average build time for a civilian ship is 1 week per 5Mcr (HG v1, not sure if that changed). Military ships use a separate construction time table. However, with that being said, shipyards are only going to construct ships if either there is a buyer already or an existing waiting list. The average life span for smaller traders is measured in decades. In the real world ships get replaced about every 25-30 years, mostly because new technology has come out or there has been a change in the underlying economics (such as bigger and bigger container ships that more efficiently transport their cargo). Much of that has to do with it's usually cheaper to build a new ship than to gut the first one and have to replace the engineering section. Only military ships, with their myriad other systems, make sense to do that sort of costly work. If the average life expectancy of a free trader is 80 to 100 years, that will affect just how many get built anywhere per year.

Which means most shipyards aren't going to be regularly building free traders (for example) to sell unless the local market is experiencing an uptick. Meaning you will probably have shipyards in a subsector that are idle, or are involved in doing more repair and overhaul than new construction. The replacement rate for hulls is a HUGE issue when it comes to determining just how much idle capacity is available for new builds. It means yards are going to be sized to support averages, and that also means most yards are going to work very hard to have a backlog. A player dropping a few hundred megacredits might have to wait a year before a slip open up for them to begin construction on their custom order. And this plays into the annual maintenance as well. Bigger firms are going to reserve that in advance. The average player shifting sectors may find no available space for months to do their work when it's needed.

TLR - Don't try to model the shipyard economy. It's far too complicated for a game with no actual economic model.

 

[EldritchFire]

Something that complicates the model (gee, ANOTHER thing??) is that the average build time for a civilian ship is 1 week per 5Mcr (HG v1, not sure if that changed)…

That has changed a bit:

"On average, assume that it takes one day per million credits to build a spacecraft at an average commercial shipyard. At the referee's discretion very large ships can be built in a modular fashion allowing simultaneous construction. This will mean the total construction time can be reduced by up to 90%. This is typically done only on ships exceeding 50,000 tons."

Assuming a week is 5 days, then it hasn't changed. But I think a week is 7 days, so it's a bit faster to build then in 1e.

 

[Galadrion]

I don't particularly care for basing construction times on ship price. As a rule of thumb, I set the construction time for any ship I regard as either "custom" (typically, designed by the player) or specialized (anything that comes from the shipyard armed, particularly with anything heavier than turrets, qualifies for this, as do a few of the other designs which assume customizing touches such as the lab ship or the yacht) at roughly the square root of the displacement tonnage in weeks, while more generic ships have construction times closer to the cube root of the displacement tonnage in weeks. This works pretty well, for the most part - not that this is an issue that comes up frequently in most of my games...

 

[Condottiere]

1. Modular manufacture, and then an assembly area.

2. Commercial ship buyers will prefer off the shelf components, especially engineering, probably also private ones for their yachts, though the hull, the interior layout and the fittings might be customized and lavish.

3. Military contracts might be spread out amongst a range of subcontractors, and may only be for components, but would represent a steady stream of income. Downside might be costs for increased security.

 

[EldritchFire]

I don't particularly care for basing construction times on ship price. As a rule of thumb, I set the construction time for any ship I regard as either "custom" (typically, designed by the player) or specialized (anything that comes from the shipyard armed, particularly with anything heavier than turrets, qualifies for this, as do a few of the other designs which assume customizing touches such as the lab ship or the yacht) at roughly the square root of the displacement tonnage in weeks, while more generic ships have construction times closer to the cube root of the displacement tonnage in weeks. This works pretty well, for the most part - not that this is an issue that comes up frequently in most of my games...

Basing build-time off of tonnage doesn't make much sense to me. The more you cram into a ship, the longer it should take. I mean, empty cargo space is quicker to 'install' then upgraded sensors, accommodations, weapons' turrets, etc.

The more bling on your ship, the longer I think it should take. That being said, I agree that it won't come up in most games.

 

[Galadrion]

For most ships the players are going to be interested in, basing the build time off of volume is a "good enough" solution. Players are going to be interested in either cargo/passenger ships (the cube-root case above) or a light quasi-military ship for either mercenary work or piracy/anti-piracy/privateer work (which would use either the square-root case or, if they used one of the off-the-shelf ship designs from the book and went with turret weaponry, the cube-root case).

Bear in mind that, for most adventurer-level ships (100 to, say, 1000 displacement tons), a general-purpose ship is going to take between 4.6 weeks (a month, more or less) and ten weeks, depending on size. If you feel the players are crowding in more goodies than would be justifiable in that time frame, well, that's a pretty good indication that this is a more specialized design. A specialized/custom ship (which I would definitely say includes any design using barbettes, bays, or spinal mounts, as well as less common design elements such as labs, medical bays for more than a few patients at time, manufacturing equipment, or similar gear) in the same size category would take between ten weeks and thirty-two weeks to build, certainly a reasonable estimate.

Larger ships scale similarly. At the 30,000 ton level (a large freighter for the "general" type, or a large destroyer/light cruiser for the more specialized/military type), you're looking at thirty-one weeks (roughly seven months) for the more general design, one hundred seventy-three weeks (about three and a half years) for the military/specialized design. I feel the scaling holds fairly well all the way up - a million-ton dreadnought would take roughly twenty years to build, which seems about right to me. A million-ton cargo vessel (which I have problems seeing a need for, under the Traveller economic model) would take around two years to build. To be honest, these estimates seem to be in line with your "empty cargo space is quicker to 'install' then upgraded sensors, accommodations, weapons' turrets, etc." observation.

 

[EldritchFire]

I can see the logic behind your argument, and it removes the need for the "reduce time by 90% for big ships" rule.

And I missed where the build-time is in weeks, not days like the normal rules. And that does change a lot. When dealing with a timespan that is rated in weeks, having a bit more stuff to put in doesn't require additional weeks to add—more like a differences of a few days, which is just a rounding error.

 

[Tom Kalbfus]

What if a ship is a hollow sphere? Lets say it is 100 meters in radius, but the decking is from say the Hull to 10 meters in and inside is nothing but empty space which the spaceship encloses. It could be filled with an atmosphere but no structural components? So how do you calculate the volume of such a ship for construction purposes?