Budget-Friendly Guide: Passing Level III with CFAI

Many guides here recommend the use of prep providers, which can be quite costly, on top of the hefty exam fees. So I decided to share my experience of passing Level III in the 90th percentile with CFAI materials for candidates with tight purses.

It’s also my way of thanking Analystforum for helping me to solve difficult questions. I did not have an account previously but surfed these forums whenever I encounted weird questions in past papers.

Reading The Text

Since I was working full-time, I knew I had to start preparing early in Sep / Oct. I planned to rely solely on CFAI materials from the onset, instead of Schweser (which had served me well for Levels I and II). Many posters have commented on the poor quality of Schweser, and the importance of reading CFAI text for the AM paper.

There were approximately 20 study sessions with 60 readings. I tackled 2 readings every week following the volume sequence. Since the CFAI text is so lengthy, I summarised (wrote on paper) my understanding of the material / key concepts / formulae. Most readings can be condensed to 6-8 pages (double-sided), with some topics like GIPS taking more than 10 pages.

I merely read through the blue boxes because they were fairly straightforward in most cases. After reading the text, I would attempt the EoC questions, referring back to the text if necessary. Note that there are some EoC questions which can only be found on CFAI website and not in the eBook. For difficult / poorly worded questions, you can find explanations for most of these questions on Analystforum or Reddit.

I originally expected to finish all the readings by end April, leaving me approximately 2 months to practise mocks. But due to some delays, I only managed to complete all the readings by mid-May.

Practice

With 1 month remaining, I started working on the online topic tests, writing out the calculations / answers on paper. I reset topics which I performed poorly in, and at the same time, read the relevant sections in the CFAI text again / googled to refine my understanding.

My score in these mocks generally was above 80% percentile. But I had seen some of these questions in the EoC questions before, so naturally I knew how to solve them and discounted my score.

In June, I started working on mocks, putting more emphasis on AM papers. CFAI provides the past 3 years of AM / PM papers. You can find mocks all the way back to 2008 on a certain Chinese forum.

I printed out 11 AM mocks (from 2007 to 2017); CFAI did not provide an AM mock for 2018. I also used 12 MCQ (PM) mocks from 2013 to 2018. These comprise of 1 (from 2013) + 3 (from 2014) + 2 (from 2015) + 2 (from 2016) + 2 (from 2017) + 2 (CFAI website).

I restricted the PM questions to 2013 to 2018, because the older questions are less relevant and there is little value in doing more MCQs, since you would probably have encountered many of these questions in the online topic test / EoC questions. For AM questions, I went further back because they were completely new. You can eliminate irrelevant questions by comparing the Learning Objectives in the Solutions with the Learning Objectives in the curriculum.

When I did the AM questions, I did not strictly time myself and occasionally referred to the text / Solutions to check if the questions were relevant. I used a smooth pen and was fairly confident that I could finish on time. As for grading, I merely checked my solutions against the guideline answers but did not assign marks (because of the irrelevant questions and inherent subjectivity).

After doing 3-4 AM papers, you start to get a hang of the common questions relating to IPS and behavioural finance. There are some less common questions that appear every few years, e.g. relating to trading systems (VWAP / algo) and economics calculations. If you read the guideline answers, you will realise that the phrasing does not differ much every year.

Example: The investor has a high risk tolerance because he is relatively young with a long time horizon and his investment portfolio has sufficient time to recover from losses / make up shortfalls etc.

Due to some delays, I could not complete 1-2 PM papers. But this was less of a concern since I had come across many of the questions before.

Exam

I wrote AM in pen, keeping strictly to the time / points for each question. Most of the questions were manageable but there were some which I was unsure of and I circled the question numbers so I could revisit them later. There was a particular Fixed Income question worth 8-marks which was completely foreign. I scribbled down the formula relating change in price to duration, hoping to get some points, and moved on. I finished the paper 5 minutes before time and did not change any of my answers.

For PM, time isn’t a major issue. You have plenty of time to revisit all the questions, including Ethics. However, don’t dwell too long on Ethics since it can be a crapshoot / quite tricky. Know the common formulae well, for example the H-model, number of futures contracts for hedging.

Post-Exam

I was confident of passing initially but got worried after reading the forums on how people who had prepared extensively failed. Fortunately, the scores were better than expected. I scored around 70% for AM and 80% for PM. Both scores were within the 90th percentile.

My experience is quite different from OP. The similar thing is I didn’t use any third party materials.

I procrastinated and began studying in March. My schedule was to read 5 books (from Behavorial finance to GIPS) in 10-12 weeks. I basically had no life in the weekend and studied 2-3 hours in the weekdays (I work full-time). I do the readings in the weekend and practiced the chapter end questions in the weekdays. I wrote extensive notes that filled up 3 notepads. By late May I finished reading the five books. Hah, unfortunately I decided to take a week and a half off from studying. When I got back, I realized I was screwed. Starting in June, I got three weeks of crunch time to review and only earnestly practiced the mock exams in the last 10 days. I read the Ethics book two days before exam. Luckily for me, I requested one week off work prior the exam. This definitely saved me.

I sucked at the AM mocks. My scores were around 50% due to poor time management. I always left at least one whole problem and a few subparts blank. I did get better at the end and scored 60% on the 2017 mock. Once you practiced 3-5 mocks, the type of answers for behavorial finance, private wealth and institution wealth problems were pretty much the same. These are the easiest points to get full credit (I aimed to get 80%+ on these sections). I only did one PM mock (the AM version from the website) and got 40/60.

On the exam, I left three subparts blank and BSed on the economic/asset allocation in the AM section. The PM section was pretty easy. I went over one more time and changed four of my answers from the first time.

My score was 55% in the AM and 83% (50/60) on the PM. This is about half a hair below the 90 percentile. I’m not surprised about my AM score, but PM felt like an outlier of my performance. To be fair, I did the PM mock one week before the exam, so it’s not truly reflective of my performance on the day of the exam.