00:00

Setting Baseline & Creating Forecast Using MFMod Outline

Start by setting expectations, collecting data, and outlining your forecast storyline. Follow steps to set a clean slate, add external and domestic variables, adjust fiscal data, and check endogenous variables. Iteratively refine your forecast using MFMod tools for accurate predictions.

azabarte
Download Presentation

Setting Baseline & Creating Forecast Using MFMod Outline

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. SETTING THE BASELINE/ CREATING A FORECAST USING MFMOD

  2. Outline - How to set a baseline 0. Write down your expectations and your knowledge 1. Start with a clean slate 2. Check or add the information that you have about exogenous external variables 3. Check or add the information on exogenous domestic variables, notably fiscal data 4. Possibly add data on endogenous variables 5. Check forecast, use model suggestions and add-factors to assess the quality of the forecast and do consistency tests. 6. Present and discuss forecasts. Revise accordingly. • REMEMBER in the near years you have a lot of information you can add. In the outer years the model has more “economic intelligence” to offer. Forecasting is an iterative process and you will iterate those steps multiple times before reaching the final baseline. •

  3. 0. Write down your expectations and your knowledge Before doing anything with the model, you should: • Collect the necessary data (notably high frequency) • Think of what you want your forecast to look like. What is expected to happen? What information (formal or informal) do you have? • What will be the main transmission channels?  This will help you write down your expected storyline REMEMBER the model will neither be wrong nor right. It is here to help you construct, assess, strengthen and sometimes revise this draft storyline.

  4. 1. Start with a clean slate Before you start, here are first unavoidable steps : Step 1: Check history Step 2: Set all sheets into exogenous mode (Change Mode > Exogenous Mode (all sheets)) Step 3: Set the statistical discrepancy, the change in inventories and net indirect taxes Step 4: Solve to update model suggestions and add factors (Solve) Then we are finally ready to start !

  5. 1. Start with a clean slate Step 1: Check history. • If history needs to be updated, please refer to later presentation.

  6. 1. Start with a clean slate Step 2: Set all sheets into exogenous mode. • Step 3: Set the contribution of 5 specific variables to zero : • Set the contribution of change in inventories and the statistical discrepancy to GDP to zero (both in volumes and values) in the forecast period • Set net indirect taxes to a fixed share of GDP (unless external information says otherwise) in the forecast period We do not want these variables to drive our forecast. Change the mode into exogenous (all sheets) Step 4: Solve to obtain model suggestions with add factors

  7. 1. Start with a clean slate Additional note: • Forecasting is an iterative process and you will iterate those steps multiple times before reaching the final baseline. Typically: year after year • The more into the future you forecast, the less information you will have about the economy. Therefore, the near years you have a lot of information you can add. In the outer years the model has more “economic intelligence” to offer. • You will add information “block by block”. Doing so and solving in between steps allows the model to update model suggestions and identities.

  8. 2. & 3. Adding Information We will go from most exogenous to least exogenous (most endogenous) variables, solving/forecasting one year at a time. Note: Exogenous/endogenous here refers to the nature of variables not to the mode option of the MFMod toolbar.

  9. 2. & 3. Start with most exogenous Foreign variables (e.g. US exchange rates, EU growth) Use WEO, WB or other forecasts Commodity pri1ces (e.g. oil price) Set statistical discrepancy, stocks and indirect taxes and subsidies fixed as a share of GDP For real taxes unless tax base widens. Nominal taxes are subject to changes in rates Usually coming from the budget office Set government expenditure envelope

  10. 2. Add new external information Set most exogenous (least endogenous) external variables Consider the effect of external variables (US interest rates, commodity prices, export market growth, etc.). Also, check: – Import and export prices as a function of exogenous commodity price indices (Keyfitz prices) – export volumes as a function of external demand (export market growth) Examine exchange rate assumption 1. 2. 3.

  11. 3. Add new domestic information Sense check and adjust most exogenous (least endogenous) domestic variables 1. Add fiscal information, notably on the expenditure side. 2. Check if inflation in the outlook is growing at the target rate or expected inflation? • We know that the long-run inflation rate should roughly converge with the target rate (the exception being if the central bank is not credible). • In the short term, use the output gap information to help adjust household consumption inflation (the model assumes a short run Phillips curve). 3. Adjust monetary policy assumptions

  12. 4. Checking endogenous variables: volumes Add information on endogenous variables if any • Production data on industry and agriculture • Consumption, investment, etc. forecasts • Price dynamics If no data, follow model suggestions for a first pass. Note that: • Production of services is computed as a residual

  13. 4. Checking endogenous variables: production ????= ? + ? + ? + ? − ? + ????? + Δ? = ?????+ ??? = Σ????????+ ??? Productionapproach Expenditureapproach + GDP at market prices GDP at factor cost Indirect taxes less subsidies + Agriculture Investment + + Consumption Industry Exports Services Imports

  14. 4. Checking endogenous variables : and finally imports Prices : LR = target while SR considers output gap Export market growth NIA Imports = f(GDE, PM/PC Exports = f(PX/PC, XMKT) Wage = f(inflation, productivity) Investment = f(UCK- MPK, Pot Growth) GDE Labor = f(population) Consumption = f(wages, wealth) Government investment and consumption Expenditures in the fiscal account Wage bill Remittances

  15. 4. Checking endogenous variables: values Values page contains mostly identities. • Value = Price * Volume, so that your values reflect your assumptions made elsewhere. • Exception: Change in Inventories , Statistical Discrepancies

  16. 4. Checking endogenous variables: fiscal Tax revenues With previous information, you should now have an estimate for the tax bases (GDP, imports, consumption etc.). Use the tax information (hint – calculate effective tax rates) and make adjustments using the nominal tax bases too. With expenditures and revenues set, fiscal deficit and debt are then accounting results. Remember : All real variables grow at the same rate All prices grow at the same rate All nominal variables grow at the same rate

  17. 4. Checking endogenous variables : fiscal Tax revenue Tax base Tax rate ???∗ ??= ??? ???∗ ??= ??? ???∗ ??= ??? ?????∗ ????? = ????? ???? ??????∗ ?????? ?? ?????∗ ????? ??? ???????? ?? ???−??? ???????? ?ℎ???? ?? ???∗ ??= ??? ???∗ ? ?????? ????? ?? = ???? ?????∗ ?????

  18. 4. Checking endogenous variables : fiscal Review on effective vs. statutory rate In real life: tax revenues = statutory tax base * statutory rate In the model life: tax revenues = proxy tax base * effective rate Both can be linked through the coverage ratio: tax revenues = proxy tax base * coverage ratio * statutory rate

  19. 4. Checking endogenous variables : BOP For the Balance of Payments : • Exports and imports are determined by your forecast on the NIA side, the USD/LCU exchange rate and trade prices so that the trade balance is set. – If you don’t like your BOP forecast, maybe you don’t like your NIA forecast. • On the BOP side, you can adjust: – The decomposition between goods and services in exports and imports (services are the balancing item so that you can set merchandise imports and exports) – Other current account items.

  20. 4. Checking endogenous variables : BOP As a reminder: XVAL,USD− MVAL,USD PXXVOL− PMMVOL E Trade Balance = In forecast, this holds in growth rate only = CAB = TB + Credits − Debits Primary and secondary incomes Credits = Remittances inflows + Others Debits = Remittances outflows + Gov. extl.interest payments + Others

  21. 5. Checking the forecast and adjusting • Now you are ready to “Solve” the model and have a first rough estimate for actual GDP • Does the model’s forecast look OK? Why not? Discuss inconsistencies. Remember a model only knows what it was designed to know. • Revise the model and add necessary information.

  22. 5. Consistency checks Elasticities • For most countries, the income elasticity of consumption and imports are roughly equal to 1. • Elasticity of consumption to output should be roughly = 1 • Elasticity of services to consumption should be roughly = 1

  23. 5. Consistency checks • Gaps (output gap, unemployment gap, interest rate gaps) should close in the long run • Terms of trade (PX/PM) and real exchange rate (Pc/Pc*) should tell the same story • Tax revenues should grow in line with their tax (nominal) base • Current account should move in line with the NIA trade balance

  24. 6. Present and discuss forecasts Once accordingly. Discuss and publish. forecasts are calculated, revise

More Related