Investment Banking Analysis

01Purpose and scope

• This manual supports the course Investment Banking Analysis. It describes how analysts in advisory and capital markets groups produce the models, comparable company sets, transaction screens, and presentation pages that bankers rely on when advising on mergers, acquisitions, equity offerings, and debt financings.
• The work covers public company and private target analysis, from the first data pull through the final review before material goes to a client or counterparty. It is written for analysts, associates, and the vice presidents who review their work. It does not cover trading desk analytics, equity research publication, or legal drafting of transaction documents, which follow separate procedures and specialist reviews.
• This guidance does not replace qualified professional judgment, approved firm procedures, or required licensing. Securities, disclosure, and conduct rules vary by jurisdiction and change over time, so check the current requirements with the securities regulator in the relevant country and with your firm's compliance group before relying on any practice described here.

02Core principles

• Every figure in a banking output should trace back to a source. That source might be a filed annual report, a quarterly filing, a data service pull with a date stamp, a management projection received under a confidentiality agreement, or a clearly labeled analyst assumption. A number without a source is treated as a placeholder and should be flagged as one. Reviewers at the associate and vice president level will ask where each input came from, and the answer needs to be quick to find.
• The model should be only as complicated as the question requires. A merger analysis for a strategic buyer usually needs an accretion and dilution schedule, a pro forma capital structure, and a synergy bridge. A standalone valuation for a board discussion usually needs a discounted cash flow analysis, trading comparables, and precedent transactions presented as a range. Adding tabs that no one will use makes review harder and raises the odds of an error hiding in a formula.
• Present ranges instead of single points. Valuation outputs are sensitive to growth, margins, discount rates, and terminal assumptions, so a single implied share price implies more precision than the work supports. Sensitivity tables should show which assumptions drive the answer, and the summary should say what has to be true for the value to reach the top or bottom of the range. Bankers use that framing to have a conversation with the client rather than to announce a number.

03Planning and preparation

• Start by clarifying the question with the senior banker who owns the client relationship. Find out whether the firm is advising the buy side or the sell side, what decision the client faces, what deadline applies, and whether the output will go to a board, a committee, or a counterparty. These answers determine the level of detail, the peer set, the valuation methods, and how conservative the assumptions should be. Ask who will present the output and what questions they expect to field.
• Gather inputs before building anything. Typical inputs include the last several annual and quarterly filings, earnings call transcripts, consensus estimates from a licensed data provider, a current share count from the latest filing cover page and treasury stock method inputs, and any management projections provided under a confidentiality agreement. Record the date of each data pull in an assumptions log. Market data changes daily, and a comparables table built on last week's prices will look wrong if anyone compares it to a live terminal.
• Set up the working file before data entry. Use a standard template with a cover sheet, an inputs tab, a historical financials tab, a projections tab, valuation tabs, and an output page. Agree on file naming and version control with the associate. Decide early whether the time frame is a first pass in a few hours or a full model over several days, because that choice drives everything that follows. A quick first pass can be useful for a discussion with a senior banker, but it should be labeled as preliminary and should not be sent outside the deal team.

04Recommended workflow

1. 1
   : Define the output pages and the decision each page supports. Sketch the football field, the summary of financial metrics, the accretion and dilution table, or the comparables summary before opening the model. Write one sentence under each page stating what question it answers. This prevents building analysis that the presentation never uses.
2. 2
   : Build the historical financials and normalize them. Enter revenue, operating costs, depreciation, interest, taxes, and capital expenditures from filings, then reconcile totals back to the reported statements. Identify one-time items such as restructuring charges, litigation settlements, gains on asset sales, and impairments, and show the adjustments separately so a reviewer can accept or reject each one. Check that the balance sheet balances and that the cash flow statement ties to the change in cash.
3. 3
   : Build the projections. Start with the management case if one is available, and build a banker or sponsor case alongside it where the two differ. Keep the drivers explicit, such as volume growth, price, gross margin, operating expense ratios, working capital days, and capital expenditure as a percentage of sales. State in writing where the cases differ and why. Do not overwrite management numbers without a documented reason.
4. 4
   : Build the discounted cash flow analysis. Compute unlevered free cash flow from the projections, then calculate the weighted average cost of capital using a risk-free rate, an equity risk premium, a beta from a defensible peer set, a cost of debt, and a target capital structure. Estimate terminal value using both a perpetuity growth rate and an exit multiple. Compare the two implied figures to each other. If the terminal value accounts for most of enterprise value, state that clearly and show the sensitivity.
5. 5
   : Build the trading comparables. Select peers by business mix, size, growth, margin profile, and geography, and document the rejection of any company that was considered but excluded. Calendarize financials to a common fiscal year where needed. Calculate enterprise value to revenue, enterprise value to EBITDA, and price to earnings using consistent definitions of net debt and diluted shares. Show the median and interquartile range rather than relying on the mean, which can be pulled by one outlier.
6. 6
   : Build the precedent transactions analysis. Screen deals from a data service by industry, deal size, date range, and whether the target was public. Record the announced equity value, enterprise value, and multiples, and note whether the premium or multiple reflects a control transaction. Adjust the interpretation for market conditions at the time each deal was announced, since a multiple paid during a period of high valuations may not apply to a current discussion.
7. 7
   : Assemble the summary and run the review. Combine the methods into a football field, check that the ranges match the sensitivity tables, and confirm that every page has a source note. Walk the associate through the model with a short written list of open items. The vice president then reviews the story the numbers tell, and the managing director approves the judgment behind any recommended range before the material goes to the client.

05Quality checks, risk controls and compliance

• Investment banking work often involves material non-public information, so access to deal files and models is restricted to the deal team, and information barriers between groups must be respected. Never use confidential client data in public materials, training examples, or personal files. Data license terms from market data providers also limit redistribution, and those terms should be checked before a screenshot or table is shared outside the firm. Wall crossing, conflict checks, and approval for external sharing are set by firm policy and by regulation, and the current rules should be confirmed with compliance.
• Model checks should be run before every review. Confirm that the balance sheet balances in every period, that cash flow ties to the balance sheet, and that debt schedules reconcile to interest expense. Check that formulas are consistent across each row, that hard-coded numbers inside formula ranges are flagged, and that any circularity has a switch to break it. Verify sign conventions for capital expenditure, working capital changes, and tax. Many firms use a separate analyst to run an independent check on the final version, and that person should not have built the model.
• Risk controls also apply to the outputs. Confirm that the valuation date, share price date, and currency are stated on every page. Recheck the diluted share count against the latest filing, since options, restricted stock, and convertible securities can change the answer materially. Where the deal involves cross-border companies, identify whether financials are prepared under US GAAP, IFRS, or a local standard, and document any adjustments. Accounting and disclosure requirements differ by jurisdiction, so the current authoritative standard setter and regulator should be consulted for each case.

06Collaboration and documentation

• The analyst builds and maintains the model, the associate reviews the logic and inputs, the vice president owns the client conversation and the narrative, and the managing director sets the judgment behind recommended ranges and approves what leaves the firm. Each role should know what it is responsible for signing off. Ambiguity about who checked what is a common cause of errors reaching a client, so keep a review log that records the reviewer, the version reviewed, and the comments cleared.
• Documentation should be lean but complete. A cover sheet should list the purpose, the valuation date, the version number, the author, the reviewer, and the sources. An assumptions book should record each key input, its value, its source, and the reason it was chosen. A change log should note what changed between versions and why. Handoff notes help when an analyst rolls onto another deal or when a senior banker asks for a quick update at short notice. The trade-off is time: heavy documentation slows turnaround, so aim for the minimum that lets a stranger reproduce the output.
• Communication with the team should be direct and specific. Tools such as spreadsheet software, data terminals, presentation software, and a document management system each carry their own version risk, so the team should agree on which file is the source of truth before sharing.

07Measures and continuous improvement

• Useful indicators for an analyst team include the number of review comments per model version, the hours spent on rework, the number of errors found after material has been shared with a client, the time from request to first draft, and the share of reconciliation checks passed on the first run. Track these informally on each deal and review them in a short retrospective at the end. The point is to find recurring problems, such as stale market data or inconsistent net debt definitions, and fix the template rather than blame individuals.
• Be careful with speed as a metric. If turnaround time is rewarded without a matching concern for accuracy, analysts may skip checks that matter. A balanced set of measures pairs speed with error rates and review findings. Practice also changes across contexts. Sell side mandates, buy side fairness analysis, leveraged finance, and equity capital markets each emphasize different outputs, and a template built for one should be adapted rather than copied for another.
• Continuous improvement works best when the template and the checklist are updated after each deal. Record which formulas caused problems, which data sources were unreliable, and which comparables screens produced unusable peer sets. Share those lessons with the group in a short written note. Remember that tools and standards evolve, so review the current disclosure and accounting guidance regularly rather than relying on memory.

08Common mistakes and how to prevent them

• Mismatched periods cause frequent errors. Calendarize the financials, or at least label the periods clearly and footnote the adjustment. Another common error is double counting items in net debt, such as including operating leases in debt while also deducting lease expense in EBITDA, or counting pension deficits that were already reflected in the balance sheet. Pick one definition, apply it to every company in the peer set, and document it.
• Share count errors are easy to make and hard to spot in a summary. Using basic shares instead of diluted shares, or applying the treasury stock method with the wrong share price, can shift implied equity value by a meaningful amount. Recheck the count against the latest filing and the option schedule. Terminal value dominance is another trap. When the terminal value makes up most of the enterprise value in a discounted cash flow analysis, the output rests heavily on long run assumptions, so the summary should say so and show the sensitivity.
• Hard-coded plugs hidden in formulas create silent errors that surface after a client has already seen the output. A plug used to force a balance sheet to balance may hide a real problem in the cash flow. Prevent this by banning unlabeled plugs, requiring any override to appear on the assumptions tab with a note, and having a second person scan for hard-coded numbers. Finally, stale market data undermines comparables. Refresh prices, currency rates, and consensus estimates on the day the output is finalized, and record the time stamp.

09Practical checklist

• Confirm the question, the decision it supports, the deadline, and the audience with the senior banker before building.
• Record the valuation date, share price date, currency, and data sources in the cover sheet and assumptions book.
• Reconcile historical financials to filed statements and document every normalization adjustment separately.
• Label management and banker projection cases and state in writing where they differ and why.
• Check the balance sheet, cash flow tie out, debt schedule, and formula consistency before each review.
• Confirm the diluted share count against the latest filing and apply the treasury stock method consistently.
• Define net debt the same way for every company in the peer set and footnote any lease or pension treatment.
• Show the terminal value share of enterprise value and the sensitivity of the output to discount rate and growth.
• Confirm information barrier, conflict, and external sharing approvals with compliance before anything leaves the firm.
• Log each review, the version reviewed, the comments cleared, and the open items handed to the next reviewer.

★Learning sources

• Micro Degree: Investment Banking Analysis
• Micro Degree: Accounts Payable Management
• Micro Degree: Bank Reconciliation Operations
• Micro Degree: FinTech Fundamentals