Threat Monitor Part 2: Looking over the horizon

Threat monitors (described in more depth in Part 1) are a great way at tracking of tracking events in the present and near future.

So what happens when you want to look further? If you’re concerned with longer-term risks, like investments that last five years or more, how can a threat monitor work?

The benefit of medium-term forecasting

The proper time horizon for political risk is always dependent on the question. 

The UK Ministry of Defence has projected out to 2050. Hedge funds the night before the Brexit referendum forecast less than 24 hours ahead.

Medium-term forecasting - between three and seven years away - lies in between. Methodologically speaking, it’s a challenging time period to analyze. 

On the one hand, it’s too far away to assume that present trends (or questions) will persist. If an analyst in 2010 was asked about the medium-term future of UK politics, they would have had to foresee austerity measures, a European double-dip recession, a Conservative victory in a General Election, a refugee crisis on the Continent, and the Brexit campaign. We’ve known some very good analysts, but that would have been beyond anyone’s forecasting skills.

On the other hand, the medium term is too soon to assume that an event is truly unforeseeable. Forecasts past fifteen years out are usually done as scenarios; it’s accepted that anything else is unrealistic.

Medium-term forecasts are stuck between far enough from the present to be overwhelmed by uncertainty and yet too early to be done as scenarios. There are few investments that are planned to only last a few years and few policies that we assume won't be in place that long.

Monitoring the medium-term

One way to forecast possible threats a few years’ out is a threat monitor. As in the more immediate threat monitors discussed in Part 1, we here follow the risks that emerge from a variety of sources.

The difference lies in what elements are chosen to construct the monitor. In the US-China soft power monitor, we included drivers like chaos in the 2020 presidential election. Obviously, that will not be very relevant for where the US and China will be in 2025, when that race will have receded into the past. Similarly, the impact of COVID-19 will have dissipated by then.

In a forecasting threat monitor, the construction must first start with the drivers. This could involve changing the scope of a driver. “Chaos in the US presidential election in 2020” could turn into “Partisan gridlock and political legitimacy questions in US politics.” More likely, it’ll mean a completely new or wider driver.

Then, just as with an immediate threat monitor, the drivers are weighted, scored, and added.

For our US-China soft power in 2025 question, our monitor looks like this.

Utility of a monitoring a forecast

As with a near-term threat monitor, the real utility comes when it is revisited. By updating the monitor, you’ll see whether the threat is rising or falling. Quarterly or semi-annual updates are best for medium-term threats.

By updating the monitor, you’ll see whether the threat is rising or falling. Regular updates act as a convenient organizational hook on which to hang analysis that can get lost in the day-to-day. It’s much easier to assess a situation as part of a scheduled exercise than wait for someone to raise it to the boss’s attention - especially if it’s a low-probability event like those that the monitor is designed for.

Updating the scores and weights of the monitor also forces us to consider whether the drivers are out of date. Perhaps revising this monitor in 2021 we will see that China has succeeded in quelling international outcry about Hong Kong. In that case, it would be removed from the monitor as it’s no longer moving. Instead, if there is growing international conversation on China’s actions in Tibet, that should be included. This not only keeps the monitor reflective of the situation, but can also force a team to reflect on their priorities. If the US-China soft power monitor were run by the State Department, maybe that realization forces leadership to put more weight on Tibet. Maybe it causes them to raise the issue of Hong Kong again because they are reminded that it’s fallen off the international agenda.

Revisiting a medium-term forecast after three or six months accomplishes one of the most important yet neglected tasks in political analysis. It forces us to look back at our old predictions and see what we got right or wrong. 

Keeping up with daily events can distort our thinking in two ways. One, it can turn us into the frog in boiling water. We don’t know how much has changed because we see it moving too slowly. Two, it can turn us into a headless chicken. We chase after every small move and we don’t realize that nothing significant has happened in the past quarter. Threat monitors act as intellectual stabilizers, smoothing our predictive movement, and turning us into better forecasters. 

There is one additional benefit to threat monitors. By adding uncertainty measures to each of our drivers, we can use Monte Carlo simulations to approximate risk potentials.

We will cover that in Part 3.