it came from the data (UN Comtrade) that notably doesn't include sanctioned oil flows (Iran, etc). On top of that, China's reserve levels are a state secret, yet, the model's framework is BYOD (Bring Your Own Data) which gives the depletion timeline for free. Also, the model is more of a sensible baseline than a prediction
Thanks! Two answers (repeated in other comments): UN Comtrade data doesn't include sanctioned oil (Iran), and nobody (except their government) knows China's actual stockpile. The model is more of a sensible baseline/stress testing tool than a prediction, the cool part is that anyone can BTON (bring their own numbers) and rerun their scenarios
the models, proofs and writing are mine (with some AI assistance for proofreading/formatting). AI was mostly used in the visualization and the rerunning of multiple scenarios of the numerical examples
This pipeline predates the data but its oil is routed to Bab-el-Mandeb which is in the network. That being said, the numbers are more of a sensible baseline as the actual flows for many countries are unknown/sanctioned.
France imports 0 oil directly through Hormuz, but when the Strait closes, all other countries start raising their safety stock which increases oil prices. This makes France's stockouts very expensive mostly due to price (not flows). Which is contagion without direct connection.
Think of the 2008 crisis where companies that had nothing to do with Credit Default Swaps were exposed to the crash due to network effects, potentially crashing the whole economy.
Then, the government intervened by bailing the banks out. Here, it's not really clear if there is even an effective intervention.
this is an interesting market design question: the point is that tapering would require everyone knowing everyone else's depletion clocks. As they are unknown (sanctioned trade, non-public state figures), the opacity is the game itself (that I called game of chicken in other comments). Surely the main players are trying hard to guess each other's numbers by every mean possible
That's actually the intended behavior: while reserves are silently being spend is when the crisis is brewing. Although the main point is that nobody knows China's reserves, but this allows to calibrates scenarios based on what you believe is plausible
And this is what the model price formation assumes, and in fact, the silent mechanism that makes the crisis worse. Reserves silently deplete for each country, and each epoch where they exhaust is when the price rebalancing occurs IN A SUDDEN SPIKE, affecting other nodes that are not even connected to Hormuz, which is one of the conclusions of the paper. Either directly or indirectly all countries feel the pain: the question is who can stand the game of chicken the longer before intervening
There are two versions of the model:
-The one rendered by default (called Endogenous), with endogenous price that affects the demands according to the price elasticity slider. Reflecting how reactive are the nodes demand wrt price
-With fix price (called Fixed): that in the oil market doesn't apply, but it's an interesting baseline to consider oil purely as flow. And as a modeling tool, this allows the model to capture small supply chains that don't have price setting power.
Thank you. Yes, in desktop there is one (that might still be going a little bit too fast), but in mobile it might not be that easy to find. Perhaps I will slow down the speed in mobile by default and change the location (should be in the bottom right corner, the 2 wk/s can be changed to 1 wk/s)
Fair point. Actually this is both a semantics oversight on my part and also expected behaviour. at 100% retained there's no shock, but the model's countries target consumption plus a large safety buffer, and some can't fill that target even in peacetime, so they draw reserves at baseline, the shock scenarios show the additional damage relative to that baseline
This is a nice idea. I will build this a feature on the scenarios menu. And it would give an estimate of how much time doing this buys China under different configurations