Cash Flow Forecasting: Creating a Solid Foundation Ken Parkinson Treasury Info. Services tisconsulting.com & tisbooks.com © Treasury Information Services, LLC 2011 All rights reserved. 1 Forecasting Setting Investors (companies) pay [far] less attention to forecasting than they would if they borrowed. Why is this usually the case? So what to do if you’re an investor? How is your cash forecast used or regarded? By the sources of data By management users Does anyone seem to care? © Treasury Information Services, LLC 2011 All rights reserved. 2 Forecasting Variables 1. 2. 3. 4. 5. 6. 7. Business maturity: more mature, easier to forecast Historical predictability: similar to previous Cash flow decentralization: hard to get data, but … Forecasting period: smaller ones - more accuracy Accounting influence: the less, the better Ease of data collection: harder to get – more errors Org. level of source: too high, bad estimate © Treasury Information Services, LLC 2011 All rights reserved. 3 The Forecasting Cycle Sources estimate & send data Data compiled/r eviewed Variances computed & fed back to sources Cash flow forecast prepared Actuals compared to forecast © Treasury Information Services, LLC 2011 All rights reserved. 4 Forecast Horizons Daily – rolling 5-days; a necessary part Weekly – rolling 4-6 weeks; often ignored Monthly – rolling 12 months; a “staple” Quarterly – rolling 4-6 quarters; necessary? Annual – rolling 5 years; a different context © Treasury Information Services, LLC 2011 All rights reserved. 5 The Problem with Forecasting How your boss sees it – “Just push the button.” The actual nature of your cash flow forecasting “challenge” How do you reconcile this? © Treasury Information Services, LLC 2011 All rights reserved. 6 So Much Unhappiness! 75-80% lf senior execs do not consider their cash flow forecasts reliable. Why is this? Exactly what do they dislike? Do firms evaluate forecasting properly? Or do they try quick fixes? Or do they give up? Why don’t many cash flow forecasting systems work? © Treasury Information Services, LLC 2011 All rights reserved. 7 Managing Expectations Whose forecast is it? What is in your forecast? Compilation of data Best guess Worst case What are you forecasting? Senior management? Sources? How about – YOURS? Cash position Cash flow (and cash balances) Why are you forecasting? For senior management For strategic planning, esp. financing © Treasury Information Services, LLC 2011 All rights reserved. 8 Managing Expectations How often do you need to forecast? How much do you already know? Daily? Weekly? Monthly? Annually? How often can you forecast? Different ways to forecast? Financial transactions? Large operational cash flow items? Repetitive customer (i.e., “regulars”) payments How much don’t you know? Surprises Variances Do you rate the sources? © Treasury Information Services, LLC 2011 All rights reserved. 9 What Bad Forecasts Do to You Create severe constraints on working capital. Shorten investment maturities ( = lower yields) or causes investment liquidity risks (bad guesses). Increase inefficient short-term borrowing Overuse credit facilities. Lines that are never used. Create excess fund situations in local accounts. Cause inefficient planning for working capital management and long-term financing. Lost opportunities and/or more exposure to ST rate changes © Treasury Information Services, LLC 2011 All rights reserved. 10 What’s the Best Structure? Should mirror organization’s cash flows Benefits of decentralized structure Centralized vs. decentralized Can’t ignore potential data sources Many data points reduce dependence on few Isolate problems at operating levels Possible benefit from offsetting errors Benefits of centralized structure Fewer points mean easier data gathering Better control over data, submissions Ignore local politics © Treasury Information Services, LLC 2011 All rights reserved. 11 Which Structure? Top-down Work from financials What’s the “top” for daily-weekly-monthly? Probably better for LT models; maybe monthly May be statistical model Bottom-up Nuts + bolts; no statistical models More [real] cash flow oriented Perhaps easier to grasp and use © Treasury Information Services, LLC 2011 All rights reserved. 12 Top-Down & Bottom-Up Forecasts Top Down: Start with sales Receivables, payables, payroll, inventory, cash all can be related to sales Use historical % of sales to project, based on a good sales forecast © Treasury Information Services, LLC 2011 All rights reserved. Bottom Up: Start with detailed projections Get estimates of components (receipts, disbursements, etc.) Roll up into total forecast Which is better? 13 Forecasting Approaches Top down – for medium-term (i.e., monthly and up) Bottom up – for everything up to annual Is it either-or … or both? The case for either: simplified system; pick most accurate, cost-effective The case for both – redundancy can catch errors Forecasts should be ‘rolling’ © Treasury Information Services, LLC 2011 All rights reserved. 14 Cash Flow Forecasting System © Treasury Information Services, LLC 2011 All rights reserved. Source: Cash Flow Forecasting: A Hands-on Approach 15 Integrate Your Forecasting Shorter forecasts link to longer ones Daily should reconcile with the 1st week of the weekly Weekly should reconcile with 1st month of monthly Monthly should reconcile with 1st year of annual © Treasury Information Services, LLC 2011 All rights reserved. 16 Types of Sources People Other company systems Managers, et al. close to activity Should be “experts” in expert judgment systems Key advantage: Minimal human intervention ERP systems Transaction-oriented systems Independent estimates Financial models, etc. © Treasury Information Services, LLC 2011 All rights reserved. 17 Operating and Financial Flows Operating Flows Possibly many sources = hard to get Recurring payments means history useful New, identified flows Planned budget can give some idea Base levels may be relatively easy to estimate, reducing difficulty Financial Flows Reactive flows, depending on operating flows Often has lead time Fewer sources History can be guide, such as rollover levels LT planning can provide estimates Do you try to estimate each one separately? Why or why not? © Treasury Information Services, LLC 2011 All rights reserved. 18 History and Predictability History can be lifesaver Use in emergencies Learn from it Predict patterns from it Some things are easy to predict Some are not so easy Payroll, taxes, major cash flows, debt levels Payment timing, capex flows, inventory usage You won’t be able to predict everything. © Treasury Information Services, LLC 2011 All rights reserved. 19 Whole > ∑Parts? Unit Estimate Actuals Var. (Act-Est) % Var. Division A 6,250 4,025 (2,225) (55.3)% Division B 2,350 3,575 1,225 34.3% Division C 4.560 2,995 (1,565) (52.3)% Sub X 1.150 2,275 1,125 49.4% Sub Y 7,540 8,735 1,195 13.6% TOTAL 21,850 21,605 (245) (1.1%) Note! © Treasury Information Services, LLC 2011 All rights reserved. 20 Techniques Don’t expect to use many “sophisticated” techniques. Think this way Simple = better? “Heavier” ones don’t work because … Short-term (up to 1 yr): data manipulation (management) Longer-term (1 yr+): more ‘real’ forecasting Most statistical forecasting techniques are tough. Lots (and lots) of clean data Predictability of at least one main factor (usually sales) Regular “maintenance” and re-testing © Treasury Information Services, LLC 2011 All rights reserved. 21 Methods to Consider Method Uses Comments Qualitative classifications Short-term estimates (day to mon) Quantitative classifications Longer-term forecasts (1 yr+) Causal models Short-term projections (up to 1 yr) © Treasury Information Services, LLC 2011 All rights reserved. Expert judgment (Guesstimates) “Naive" forecasting Distribution method How accurate? Decomposition models %-of-sales – future based on past • Tough to create good models • • • • • Economic models Future based on outside factor Good for established cash flows 22 Application of Distribution Method Receivables or payables projection models Use history to predict clearance factors Influenced by Pareto’s 80/20 rule AR or AP ageing “clearance” factors applied Estimates by period © Treasury Information Services, LLC 2011 All rights reserved. 23 Analyzing Roll-over Rates Track your liquidity “balances” regularly. A spreadsheet will do. Go back and find history if necessary. Compute the average roll-over rates by month or most reliable period for ST investments and debt separately (if both present). Look for changes in net position (indicators of cash flow movements. Are they recurring? How predictable are they? Look for the minimum and maximum. Incorporate into your forecasting system and predict usage of lines or potential excess cash situations. Track continuously. Monitor and modify as necessary. © Treasury Information Services, LLC 2011 All rights reserved. 24 Roll-Over Rates max $M min Time min © Treasury Information Services, LLC 2011 All rights reserved. max min 25 Per cent of Sales CASH FORECASTING WITH PRO FORMA FINANCIAL STATEMENTS Assumed Fixed Assumptions: Depreciation (% of F/A) Div idends Notes Will Be Inc/(Dec) f or Next Y ear by Interest Rates on: Notes Bonds Projected Sales ($) 5.00% 6.75% $2,000 Variable Percentages: COGS as % of sales Selling/Admin Exp Cash as % of Sales A/R as % of Sales A/P as % of Sales Inv entory as % of Sales Taxes Additions to Fixed Assets ($) Cash Accounts Receiv able Inv entory Net Fixed Assets Total Assets Actual $250 300 200 1,000 $1,750 Projected $333 400 267 1,045 $2,045 Accounts Pay able Notes Bonds Common Equity (incl R/E) Tot Liab & Equity $350 100 450 850 $1,750 $467 100 450 996 $2,012 Actual $1,500 (850) (275) (100) (10) 265 106 $371 Projected $2,000 (1,133) (367) (55) (35) 410 (164) $246 5.00% 100 56.7% 18.3% (Enter % to 16.7% over-ride 20.0% computed 23.3% values.) 13.3% 40.0% 100 %-of-sales BALANCE SHEET Projected financials INCOME STATEMENT Sales COGS Selling & Admin Exp Depreciation Interest Exp Income Bef Taxes Taxes Net Income Bef ore Div s © Treasury Information Services, LLC 2011 All rights reserved. Difference $33 New debt “Freeze” Notes to compute new debt figure 26 Master Maturity Schedule Principal only shown ($000) By Type 05/11/11 05/12/11 05/13/11 05/14/11 05/15/11 Mon Tues Weds Thur Fri 500 1,000 1,500 2,000 2,500 1,000 1,500 500 1,000 2,000 2,000 1,000 Current Beyond Week 1 Week 2 Week 3 Week 4 ST INVESTMENTS Repo CP-01 CD-01 CP-02 CD-02 CP-03 750 1,500 2,500 500 2,000 1,000 1,000 2,000 1,000 4,250 3,500 8,500 5,750 5,000 1,000 2,500 1,750 1,500 500 500 2,500 2,500 3,000 2,000 5,500 5,300 4,900 5,000 3,500 Details (actuals) ST BORROWING CP Bank-01 line Bank-02 line 500 450 420 5,000 2,500 2,500 1,300 1,000 1,000 1,200 750 750 2,500 10,500 1,500 6,200 1,500 6,170 2,500 1,000 1,000 2,800 1,300 1,300 1,000 750 750 10,200 7,000 7,000 5,675 4,475 6,625 2,200 1,450 1,250 5,250 5,120 7,500 26,500 7,300 20,345 1,250 1,200 2,500 2,400 3,600 3,500 85,600 83,460 2,500 4,000 2,500 5,000 2,500 7,500 1,370 5,000 10,000 4,000 3,300 FX Proceeds (sales) Purchases LT TRANSACTIONS New /rolled over LTD Stock divi pmts 4,250 LTD Repay ments links 0 13,250 7,500 50,000 12,750 Summaries (links) *** Totals ST INVESTMENTS ST BORROWING 5,000 2,700 6,500 28,000 5,500 22,870 © Treasury Information Services, LLC 2011 All rights reserved. 6,500 4,500 5,500 5,400 5,500 2,500 24,200 24,200 27 Variances Measuring is not enough: You need reasons. Typically measure latest estimate (day, week, month) Measure several variances Should be best one Why is this insufficient? Several intervals Gives you a better picture How to “fix” variances Communications Adjustments to current and/or future forecasts © Treasury Information Services, LLC 2011 All rights reserved. 28 Charting Variances Note change in direction. Net cash flow for 4 mos. $ Note shift here. © Treasury Information Services, LLC 2011 All rights reserved. 29 Variances for Intervals Variances Net cash flows 1-mo 3-mo 6-mo 1,170 1,600 (400) (680) 700 1,100 400 200 900 3,600 7,600 (900) (2,700) (4,600) 500 1,500 4,000 1,500 4,500 2,700 (300) 700 2,300 200 600 5,900 200 600 (3,400) 731 1,818 4,357 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Avg Avg 1.75% 4.34% 10.40% %-Var © Treasury Information Services, LLC 2011 All rights reserved. 30 Variance vs. History %-VARIANCE FROM AVERAGE (Est-Avg)/(Avg) Jan Receipts Staff deposits 24.1% Chicago lockbox -37.5% Atlanta lockbox -16.7% Maturing investments -41.5% Customer ACH payments -6.3% Miscellaneous cash inflows 3.6% International wires 60.0% ** TOTAL RECEIPTS -14.9% Feb Mar -17.2% 24.1% -43.8% -56.3% 0.0% 16.7% -12.2% -41.5% -25.0% 12.5% 9.6% 21.8% 140.0% -100.0% -14.5% -9.7% Apr May Jun 65.5% -25.0% 0.0% -41.5% -25.0% 3.6% -20.0% -9.3% 24.1% -6.3% -16.7% -12.2% 12.5% -8.6% 60.0% -4.9% -17.2% 0.0% -33.3% 46.3% -6.3% -8.6% 60.0% -4.9% 68.0% -14.3% 0.0% -34.1% -100.0% 18.2% -17.2% 11.0% 8.0% 71.4% 0.0% -7.7% -100.0% 36.4% 65.5% 21.8% 39.4% 69.4% Disbursements Controlled disbursements 20.0% 8.0% 44.0% 32.0% Payroll: checks 71.4% -14.3% -14.3% -14.3% Payoll (checks) 0.0% 0.0% 0.0% 0.0% ST debt maturing -34.1% -7.7% 18.7% -7.7% LT debt repayment -100.0% -100.0% -100.0% -100.0% Miscellaneous cash outflows -54.5% -36.4% -18.2% 0.0% Wires (outgoing) -17.2% 24.1% 65.5% -58.6% ** TOTAL DISBURSEMENTS -44.2% -31.1% -11.6% -3.4% *** NET CASH FLOW -96.4% © Treasury Information Services, LLC 2011 All rights reserved. -60.7% -14.9% 7.2% 31 Conclusions Many items may lower your overall error rate, but … How are you doing individually? What do you do? Once you receive forecasts from sources – the forecast becomes yours Act accordingly Watch out for statistical methods. They are suited to LT forecasts. You won't be able to predict everything. But you want to be in the ball park. And remember -- it’s just a forecast! © Treasury Information Services, LLC 2011 All rights reserved. 32 Any Questions? © Treasury Information Services, LLC 2011 All rights reserved. 33