Tuesday, January 31, 2017

Personal opinion: The gloom that is the shipping industry

We have all heard about the shipping industry, how freight rates are dropping, and ships are waiting without cargo. Having been in the navy, I have actually seen those ships anchored out in open waters (so that they don't have to pay charges for being in harbour) as they await instructions on where to go.

There really is a lot of excess cargo space.

Simple demand and supply tells us this: if supply is constant, and demand goes up, price increases. If demand goes down, price decreases. And in the shipping market, where it takes a while for supply to change (you can't build a ship overnight), we can safely assume that supply in the short term is a constant. Which means prices will and do fluctuate based on demand, which changes a lot more quickly based on the market needs in different countries.

But that doesn't mean supply doesn't change. It does. When prices go up, shipping companies will want to reap themselves more profits, and so they try to increase their supply of cargo space. Simple math: if I have 200 tons of cargo space being sold at $1000 per ton, and I see that demand is rising, I will want to increase my cargo space, maybe to 300 tons, so that I can earn $300,000 instead of $200,000.

But the problem is, building ships is a long term investment, and there is a life cycle cost to it. So shipping companies usually rely not just on their own ships to meet the market's demands, but also lease additional ships when necessary to meet high demand periods. The good thing about leasing is that the shipping companies do not have to bear the life cycle cost. I mean, it is the same thing as leasing office equipment. You just pay for what you need.

However, things like ships don't disappear just because they are not being leased. They hang around, and for the case of ships, they hang around for quite a while (like 10 to 15 years, if not more). So when global demand for sea freight is high, and shipping companies build ships, and others build ships to lease to shipping companies, the overall supply of cargo space in the world goes up. Which is necessary since that much cargo space is needed to move goods around at that point in time. But global demand does not always remain high. And when demand drops, we are left with ships lying around.

This is not a problem if demand doesn't fluctuate that much. Which is usually the case, since no one doubles or triples their demand for food or oil overnight. Right? Actually, no. There was actually a sudden and huge increase for sea freight demand in the mid 2000s. This drove up prices so high, it attracted many many many people into the shipping market. People started to build a lot more ships, thinking that the demand will continue and hoping to reap profits out of it.

And so when demand suddenly dropped, we were left with many ships (and lots of cargo space) on hand, and many more ships being built in the shipyards (and these ships, when completed, will further add to the cargo space supply, more than those ships being retired and scrapped). And when there is excess supply, with demand more or less the same, prices drop. Simple microeconomics.

And prices may drop so low it goes below operating cost, which means shipping companies are operating at a loss. Those without huge reserves or other forms of income will soon go bankrupt, like what actually happened to some already.

It is not an easy problem to solve. Simple thinking tells us that we have excess supply, which drives down prices. So why can't we just artificially limit the supply? Push down the supply to a level such that it drives prices up to an acceptable level to break even with the operating cost. Simple, right?

Wrong.

One is the practical part. Unlike a cup at home, where you can keep in storage when not in use, ships can't just be stored like that. They need to be regularly maintained even if not in use, else they will rust away (literally). And even finding space to store them is an issue. If you keep them in harbour, you incur wharfage. If you keep them out at sea, you need to put people on them (so that the ship doesn't drift to unknown waters, or worse, drift into something and then you end up having to pay for damages). So there is a cost to keep them lying around.

And no one wants to drastically reduce the number of ships they have. After all, if demand does go up again, they want to have the capacity to meet that demand, and it takes time to build a ship.

Even if one company artificially tries to limit its supply, there is no stopping its competitors from doing otherwise. I mean, let's say a group of shipping companies sat down, and decided, "Okay, let's all sell at $1000." There is nothing stopping one of them (or all of them) from selling at $900 (or less) just so that they can secure more orders. And they can all do this because they have the excess cargo space to take on those additional orders.

So this is what we end up with. We have many ships, no one wants to get rid of them, and no one trusts each other enough to try and sell at a higher price. So until the supply starts to reduce (over time, as ships get scrapped), prices are going to depend mostly (if not fully) on the demand, and with demand as it is now and in the foreseeable future, I think the current situation of low shipping rates will continue for quite a while more. Analysts, especially those in the shipping industry, try to see things in a more complicated, and optimistic, way, but the simple truth is there are too many ships, and no one wants or dares to do anything about them. And since no one likes to bring bad news to the boss, there will always be people trying to give optimistic predictions. But as an outsider, my take is that it will be a long while before prices go up, and good luck to all ship owners until then.

(Why we ended up with so many ships is another issue altogether, brought about by non-traditional players entering the shipping market when prices soared. But it will take another article to fully talk about that issue.)

Monday, January 30, 2017

Where is meritocracy going?

Trump is closing doors to the US for certain groups of immigrants. And adopting economic policies that will put Americans first (which I think is something that every country does, they put their citizens first), but in a way that may limit competition.

If you ask me, competition is the way human beings improve. Fair competition is also what allows meritocracy to survive. If people are not allowed to compete on fair terms, the best will not be rewarded; we are instead rewarding those who may not be as talented or as hardworking.

If we believe that people should be rewarded for how hard they work, and how much they contribute to society, then we need to lower or remove any form of unfair assistance to certain groups of people.

In the long run, by sheltering the American people against fair competition, is Trump helping them to grow, or is he helping them to stay complacent within their own comfort zones? And ultimately stagnating growth?

How do we balance a certain level of assistance so that everyone can survive at a certain acceptable level, yet still provide room and incentive for competition, so that people continue to want to improve?

Friday, January 20, 2017

A common set of experiences

For an education or training institution to be truly great, it must create a set of common experiences that spans generations. So that every generation that passes through the institution has a common set of experiences that they can relate to, which binds them to graduates before and after them, and not just those from the same batch.

This is the binding glue that makes an institution beyond networking with the people who attended it with you. The same common set of experiences allows you to connect with those before and after, greatly expanding the network that you can build beyond your own batch.

And this is what makes a great education/training institution. Because it now is just a place of learning, but a place where you can build strong ties with not just those who attend the course with you, but with all those who have gone before you, and all those who will come after you.

That is the true power of an alumni.

Thursday, January 19, 2017

Getting to know TensorFlow

I have been trying to learn TensorFlow. But just trying to get the tutorial examples to work is killing me. Most of the examples on TensorFlow's GitHub page just can't work out of the box. It seems that TensorFlow's API is changing so fast, with methods being moved from module to module, that slightly older examples just can't work with the newer version of the API.

It is frustrating, trying to figure out what has been moved to contrib, what has been moved to legacy, what has been moved to deprecated, etc. Then having to change the source code, try running, get errors, change source code again, and run again, and the cycle continues. Until it finally runs for a while, you spend an hour reading and preparing the large data file, and then it gives another error.

Especially since the data for deep learning is usually very big. Huge.

It is a terrifying experience.

But I think it is also can't be helped. TensorFlow is the bleeding edge, and changes daily. Just that there should be a bit more organization, a bit more planning out of the API, so that we can avoid a bit more heart pain.

Wednesday, January 18, 2017

Just for laughs: TV listing on Trump's inauguration

The Sunday Herald is really humorous to put this up as part of the TV listing.

I guess we need to find humor when we can. :)

New start, again!

A first step towards another new start.

Hopefully, this will work out well! Wish me luck!

Saturday, January 14, 2017

Playing with VNC

After reading about Universe from OpenAI, I got the idea of using VNC for something similar to what Universe is trying to achieve. Basically, I need a way to pipe the output of a screen into a program, and was wondering how to go about capturing the screen. Then I happened upon Universe, and read in the blog post by OpenAI about using VNC. Which is a good idea, and so the past few days were spent trying to program a VNC client in Python. I found this and tried to adapt it to work with Kivy, which is a GUI framework for Python that I have been learning recently. After two days, I gave up trying to adapt the PyGame code in the original program into Kivy, and managed to write a Python program with a Kivy base, that can then call up the VNC client running in PyGame.

The next step is to work on a VNC server that can pipe out just the contents of the window that I want, instead of the entire desktop screen. Time to get down to learning more about the VNC framework... ;)

Friday, January 13, 2017

Open offices are bad?

Just saw this today on BBC News.
Why open offices are bad for us

A while ago, open offices, together with hot desking, was all the boom, with all the hype about creativity and such. But as with all things, there are two sides to the coin. While there may be advantages to having an open office concept, there are disadvantages too. Your company is not going to soar just because you did away with the cubicles and office walls.

In Japan, offices are usually open, without cubicles. It seems to be a Japanese business practice. And there was probably some good reason for adopting this style back in the good old days. But with the changing times and a different generation of people using a different set of technology, maybe it is time Japanese offices adopt a different layout, just to see how things may change. Some food for thought.