Advices to improve your Forecast (part 2)

Forecasting our future is built into our psyches because we will soon have to manage that future. We have no choice.
Alan Greenspan

At Steering Bird, online advisors in business management and direction, we continue talking about the forecast, one of the important activities in the finance area. In the first part, we consider general definitions. This time we will talk about some methods to do it and we will give you some advices or tips to improve yours.

Main methods to forecasting

Wall Street indexes predicted nine out of the last five recessions!
Paul Samuelson

There are several methods to make a forecast. The most common are:

  • Informal models or guessing. They depend solely on luck. Guessing is not a useful method, even if at some point, you do manage to forecast accurately. The success of a forecast based on luck would be comparable to the success of a monkey playing darts.
  • Extrapolation. Is a useful method, as long as the extrapolated trends persist, which can fall into the realm of doubt. The downside is that extrapolation cannot predict changes in trends.
  • Leading indicators. They are only reliable when they have clear reasons to predict, for example order indicators are very useful in predicting production. They lack the reliability to capture external changes, such as changes in the economy.
  • Surveys of consumer and supplier. They can be informative about future events. However, they are based on the planning of the respondents, so they provide biased information that must be analyzed before feeding any other predictive method.
  • Econometric Models and Time Series. Econometric systems, in general, consolidate existing empirical and theoretical knowledge, extracting trends from historical data or from data collected both internally and externally. Time series models rely on historical data, ordered in time (hence the name) to predict trends. They are fairly accurate, although expensive, methods. However, they can predict trends in the data quite well.

Finally, as David F. Hendry and Neil R. Ericsson point out in the book “Understanding Economic Forecasts” (2003), the success of a forecast, regardless of how it is done, requires that:

  1. “There are regularities to be captured
  2. The regularities are informative about the future
  3. The proposed method captures those regularities, and yet
  4. It excludes non-regularities”

That said, our first tip is: make sure your forecast complies with the above and you will already be taking an important step to improve the accuracy of the forecast.

How could you determine a good forecast?

It is far better to foresee even without certainty than not to foresee at all.
Henri Poincaré

Generally speaking, there is no single or standard measure to determine the correctness of a forecast. So, what makes a good forecast? In principle, the answer is accuracy and precision.

However, the accuracy of a forecast and its precision represent different dimensions of the same process. Hendry and Ericsson (2003) illustrate them with the following: “Precision almost always denotes “with little uncertainty”: one can say that the moon is exactly 5,000 miles from de Earth and be very precise, and also very inaccurate. Conversely, it is accurate to say that the moon lies between 1,000 and 1,000,000 miles away but this statement is also very imprecise.”

So, does a good forecast have to be precise or exact? To measure this, two concepts are adopted:

  1. Unbiasedness: the forecast must be focuses on results; and
  2. Variance: only a small range of results is compatible with the forecast.

For measuring projections, in statistics, there is an indicator called mean squared error (MSE) or mean squared deviation (MSD), which measures the average of the squared errors of an estimator.

Hendry and Ericsson (2003) also refer to this as mean squared forecast error (MSFE) and can be used to make forecast comparisons. It is a combination of bias and variance. However, it is not a conclusive indicator. Well, as Hendry and Ericsson (2003) indicate: “a stockbroker probably does not care how good or bad a model is on MSFE it the model is the best available for making money!”

So, we return to our question, what determines a good forecast? Depending on the context in which you work, it should be the best combination of accuracy and precision.

For example, a project forecast must be precise and accurate enough to anticipate project costs and revenues over time. Likewise, a financial forecast must have the best combination between both factors, so that the best decisions are made regarding the future of the company.

Tips for making a forecast

Prophesy is a good line of business, but it is full of risks.
Mark Twain

Finally, we give you some very simple tips for making a forecast.

  1. Use multiple scenarios. There is a tendency to be very optimistic when generating any projection or estimation. In return, very conservative projections are sometimes generated. Neither end is correct. You should generate at least these two scenarios and compare.
  2. Clearly identify your assumptions. Every forecast has an uncertainty component, which is based on extrapolations, uncertainties or projections (and in some cases, on guessing). Write your assumptions clearly and precisely. You should expressly indicate how they will affect your forecast.
  3. Don’t make a copy-paste of your budget. In the first part, we mention the differences between budget and forecast. We reiterate that they pursue different objectives. Although it is true that once a year you can work on some processes in parallel for both exercises, they are not the same and, sooner rather than later, they irretrievably separate.
  4. Reflect the processes in your forecast. You should record how your forecast changes through the different analysis processes. Save your work notes so that you can later make revisions and analyze the evolution of your estimates through the company’s processes, both in costs and sales.
  5. Analyze and review. Once your forecast is made, do an analysis. As we indicated, the most common is to forecast the financial statements. You can practice basic analysis on them, in order to see if the projections structurally conform to the company’s strategic plan.
  6. Try looking at your forecast “from the outside”. Once your forecast is finished, we recommend that you take a break and look at it with “the outside view”, with a critical sense and trying to review it without the paradigms of the person who created it. As Daniel Kahneman points out in “Thinking, fast and slow” (2011), sometimes we adopt a vision from within to estimate the future, that is “We focused on our specific circumstances and searched for evidence in our own experiences”.


I always avoid prophesying beforehand because it is much better to prophesy after the event has already taken place.
Winston Churchill

We have told you about forecast and its application in the management of companies. We have established the differences it has with respect to the budget, establishing the scope of the development of both processes. We have listed the most common methods used in this activity. We close by pointing out some recommendations to improve your forecast.

The forecast is a very important process, which in some organizations is taken lightly. Remember that the strategic plan of the company is associated with the forecast, as well as the decisions that can be made to achieve the objectives. Try to differentiate the Forecast from the Budget, in order to improve the accuracy of your projections.

We suggest you face your forecast process as if you were an advisor, or, contact us, we will help you to do it or review it.

We are Steering Bird, online advisers in direction, management and finance. We specialize in business analysis, control and analysis of investment projects, results analysis, processes of budget and forecast, etc.

Contact us if you need help. We invite you to learn about our services and read our articles.

Leave a Comment

Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text. captcha txt

Start typing and press Enter to search