When thinking recently about the role that liquidity plays in the financial markets, it struck me that parallels could be drawn between asset liquidity and something I will call “story liquidity”.
The comparison first came to me on a project where stories were “large” in the sense that we could never fit more than two or three into a two-week sprint. It felt wrong and restrictive in the same way that only having large denomination bank notes would be unhelpful if you wanted to buy something small.
It caused issues too in that the whole team often had to work on the same story (admittedly an issue that could have been addressed with sensible task writing). Worse though, it introduced an element of “all of nothing” into our sprints and meant that each story accumulated a large amount of building and testing increasing the delivery risks.
Anyway, to begin with I am going to make the not altogether unreasonable statement that the smaller a story is, the more liquid we can say it is.
Investopedia defines a liquid asset as follows:
An asset that can be converted into cash quickly (Investopedia)
I like this as a starting point. How about the liquidity of a story reflects the ease with which it can be converted into something of value quickly? This seems to make sense. Small stories, all other things being equal, should be easier and quicker to convert into value than larger stories. Now let’s go a bit further
In theory, prices in liquid markets should contain more information about the outlook for companies and the economy, because they reflect the combined wisdom of more individual trades
When that once in a lifetime (ha ha) signed Talking Heads album is up for auction, how can we value it accurately when they appear so rarely? On the other hand if I want to buy a bog-standard MP3 copy, I can look at multiple sources and make a comparison before deciding who is offering the best value.
Does this point hold for our story liquidity theory? Well, if we consider the value of a story to be the business value it will deliver once live, then surely we can make a hypothesis that the smaller and more liquid a story is, the more frequently we will be able to push our stories live and hence we will more frequent feedback/information from production usage. This in turn leads to improved insight on the true value of stories.
Perhaps the above point can also be extended to say that the more liquid/smaller a story is, the more accurate our sizing will be. As stories get smaller, teams will need to size more often and hence will have more past data to compare with and more experience in sizing. Our old friend Fibonacci plays a part here, we will be operating at the lower end of the series and hence in absolute terms there will be less variation in story size.
Let’s look another property of liquidity.
More accurate prices, in turn, should make everyone better off by improving the allocation of capital
The above statement is a common one when thinking about pricing and value. Extending our comparison once more, it doesn’t take too much effort to work out that the better we understand value and cost, the better choices we can make about how to allocate our capital – in our agile world this translates to allocation of people, time, budgets.
There are, however, two open points in this comparison that I have not figured out yet. If indeed there is anything to figure out.
Firstly, it is common in liquidity definitions to go further than I have here and say that the for a liquid asset, conversion into cash/value does not damage the price. Perhaps the mapping to agile here is to say that the delivery of a liquid story does not damage value of future stories?
Secondly, some argue that increased liquidity is good up to a point, but beyond that it can be damaging (see the link at the foot of this blog as one example). I think this is a little simpler to relate to our stories. Smaller stories are generally good, but is there a point at which where the transaction costs related to each story (maybe the release overheads?) outweigh the value delivered?