|
Post by tammio on Jun 14, 2016 18:57:16 GMT
So... some time after reaching the industrial age the societies existing on the players planet will form a single integrated economic market (looong before bowing to the eternal wisdom of the players Politburo/Empire/Union/Whatever). Actualy I beleave this is basicly a necessity of industrial age, maybe even a prerequisite. So regarding my question: Will long term market fluctuations be a thing? So basicly current economic theory teaches that free markets fluctuate in a "boom and bust" cycle with boom times (YAAY 2006) and recessions (2008), depressions (=bust; 2010), and expansions (NOW). So these short term fluctuations take 10-15 years, and are relatively weak. Furthermore theory dictates that Economies fluctuate in Kondratiev cylces of boom and bust in 45-60 years, with SERIOUS consequences.
This could be a thing in Thrive! We could "model" depressions by influencing civil unrest, income through taxes, trade (affecting taxes again) and prices. Basicly this would mean you have short term fluctuations were the factors mentioned above are changed by 1-5 units. And you have serious Kondratiev fluctuations were factors are affected by 15-20 units. This would work something like this: Boom: prices stable but high (lots of consumption), civil unrest low (low unemployment, lots of money etc), income throuh taxes high, lots of trade (meaning prices for imported goods are relatively low compared to home production) Recession: prices high but falling (deflation), civil unrest rising as recession worsens, income through taxes falls, less trade Depression: Low prices, lots of civil unrest, low tax income, little trade (meaning goods imported in one city will be expensive compared to home production) Expansion: prices low but rising (inflation), civil unrest decreasing, tax income rising, more trade. This would then affect your decissions, since during depression little income= little money left for investment, etc but lots of civil unrest means your people need something distract them (ever wondered why depressed/poor countries fight so much?) While in boom times, people are happy don't want to fight, but please invest into railways and airports and so on so....seems you anglophones name it prosperity, recession, depression and improvements. But its the same
|
|
|
Post by mitobox on Jun 15, 2016 10:41:54 GMT
I wonder if it could be possible for this to "accidentally" make it knto the game as a result of economics mechanics, like how a simulator with both tides and erosion would create shoreline beaches?
If it doesn't become a mechanic all its own, it not showing up through in-place processes could be a sign that the simulation could be improved (as in, it doesn't just happen without a cause, but those causes lead to it to show depth).
|
|
|
Post by tjwhale on Jun 15, 2016 11:59:38 GMT
Firstly the general idea of economic cycles is a cool one and you explain it really well. Secondly this comment basically nails it. I wonder if it could be possible for this to "accidentally" make it knto the game as a result of economics mechanics, like how a simulator with both tides and erosion would create shoreline beaches? The general approach with Thrive is to try and understand the underlying principles of a system, model those and then see what dynamics come out of that. So for example in predator prey systems you also get similar cycles (when there are lots of predators they hunt the prey so they are scarce so the predators start starving so the prey get a chance to rebound so there is lots of food for the predators so then there are soon lots of them etc). We're hoping that with the CPA system this kind of behaviour will just come out automatically. If we've got the relationships right then these cycles should appear. So hopefully, when it comes to the society stage, we can implement a relatively low level economic simulation which will naturally have boom and bust cycles inside it. For example a rule that would work would be "a banks willingness to lend is proportional to what it thinks the medium term growth prospects are". This means when things are going well the bank lends more and more (boom). However when things are going badly the bank will lend less and less (bust). As far as I understand it what causes things to change from boom to bust is usually companies that become dependent on debt during a boom and can't handle any contraction in lending. So basically if you need to borrow $100m to finish a building which will be worth $500m you are profitable but will still go bankrupt if you can't borrow the money. So everything is booming until a shock happens (like one large company goes bankrupt) and then the banks say "oh the medium term prospects are a bit worse, lets rein in lending a bit" but then some other companies get into trouble because they were relying on that lending and so the banks say "oh it's looking bad, stop most lending" and then you get a harsh bust. The shock of the first bankruptcy basically causes a chain reaction of tightening conditions which puts a lot of companies into trouble. Anyway hopefully we can get some nice economic models running which will have boom and bust cycles. Then you will need to chose, as the leader, what to do about them, which could be really fun. Will you Keynes or will you not?
|
|
|
Post by tammio on Jun 15, 2016 15:02:49 GMT
Well yes, I thought along those lines too, but wanted to propose a simple system for explanationary purposes first. Basicly the Kondratiev cyles I mentioned (long term cycles) are caused by Wars, technological developments, Social changes, exploitation of new Gold resources etc. Theres a quite comprhensive Wiki article about that.
|
|
|
Post by tjwhale on Jun 16, 2016 9:51:45 GMT
Hi tammio I thought what you said was interesting so I wrote a program to simulate it. Basically what happens is that there is a bank. Each time a company forms it will come and ask for a loan. If it gets it then it will slowly pay it back to the bank over time. However sometimes the company will get into trouble and ask for another loan! If the bank has the money it will give that loan but otherwise it will refuse. The company will ask 3 times and if the bank says no 3 times then the company will go bankrupt and the bank will lose the money still outstanding! Also the bank will be more willing to loan if it's been making money recently, and less willing if it's been making losses. Anyway the results were pretty interesting! I had to implement a government bailout system to keep the bank from crashing! Basically either the rate of expansion of the economy is slow enough that the bank just keeps funding and growing indefinitely. Or the bank takes on too many companies and then faces a big market crash as soon as conditions change. Here are some pictures which show the total number of companies operating in the economy. Smooth growth Harsh bankruptcy for the bank! Economic cycles with bailouts. Anyway if you want to take a look the code is here, as usual.
|
|
|
Post by tammio on Jun 16, 2016 17:24:03 GMT
Thats great! I don't know anything about coding, but I think I understood how you made it work (COMMUNISM I TELL YOU!) Since I fancy myself a disiple of F.A. Hayek I beleave it should be possible to run a simulation with positive net growth without frequent bailouts; but this is a great start. How about adding a system where successfull companies take another loan (to invest). The returns on these loans would only be 3/4 of the original rate, but the deault rate would be half(So instead of a 0.9 chance of repaying the loan is now 0.95) and the bank is more willing to grant such a loan. This might stabalise even fast paced simulations. Furthermore when the bank has only little capital reserves it would be more willing to grant these "safe" loans rather than "risky" loans.
|
|
|
Post by tjwhale on Jun 16, 2016 18:12:50 GMT
Interesting ideas. So far I haven't been able to get stable, but fluctuating, growth. Basically if the economy grows very slowly (new companies come along rarely) then the bank can just keep lending to all of them as needed. However if there is a point where the bank runs out of money, because it's the only game in town, there is a super brutal crash.
Yeah I think you are right, if the banks AI were improved (it invested with a smarter strategy) then I think things would work better.
|
|
|
Post by Moopli on Jun 16, 2016 18:37:32 GMT
Can the bank create money de novo? Or is the simulation using a fixed money supply?
|
|
|
Post by tjwhale on Jun 16, 2016 19:12:37 GMT
Currently the bank has fixed funds but the companies return the principal with interest which comes from the ether.
|
|
|
Post by tammio on Jun 16, 2016 21:24:38 GMT
Well we can have state interventions once in a while; the whole point of this exercise is to recreate economic fluctuations. All we have to do is figgure out how to stop killing the economy regularily. Once we do that we will have a working system. And a nobel prize. Maybe we could cheat as follows: In the code you created, tjwhale, companies were heavily dependant on bank loans in order to be "spawned". So once in a while a company will be created and the first thing they do is ask for a loan. Many companies do that, but not all. There are countless ways of raising capital and borrowing money is but the most boring of them. For example companies may invest in other companies just as a living and breathing person can. (The limited liability company was propably the most important invention of the Dutch EVER). WHile investment follows other rules than lending money does we can apply the same simplified mechanics here So by allowing firms to raise capital (borrow) from another firm instead of the bank the risks will be diversified in the market as a whole. Thus the Bank will not be so fast to go bankrupt (theoreticaly). Furthermore even if the bank goes bankrupt the individual companies still have access to capital. Also once a bank goes bankrupt anohter Bank could be spawned with relatively low money reserves to fill the newly created "niche" in the market. Also we could introduce an artifical "threshhold" bellow wich the Economic activity (signified in the amount of companies in the market) may never fall. We could call this our "Needs Thresshold", simbolising the economic activity that will allways happen (food still has to be baught, and cigarettes and whatnot).This means the economy will never be killed fully, nut may recover in time.
|
|
|
Post by tjwhale on Jun 17, 2016 8:37:07 GMT
tammio Have you ever considered learning to program? You seem to be really interested in this sort of stuff (economics/finance) and if you ever wanted to have a career in that field then programming would help you a huge amount. (Things like data analysis are becoming more and more standard and important). Also you could make cool prototypes for Thrive (once you were good enough). Also it's an amazing skill. Maybe it's not the kind of thing you would enjoy, but it is very powerful.
|
|
|
Post by tammio on Jun 17, 2016 12:01:22 GMT
Actualy I have considered to do so I just can't find the time. During my aprtenticeship (wich will end soon, yay!) it was more important for me to know about wine, aperitivs, drinks, cigars, etc. Maybe one day... when I'm not to occupied with my other hobbies
|
|